paperID stringlengths 36 36 | pwc_id stringlengths 8 47 | arxiv_id stringlengths 6 16 ⌀ | nips_id float64 | url_abs stringlengths 18 329 | url_pdf stringlengths 18 742 | title stringlengths 8 325 | abstract stringlengths 1 7.27k ⌀ | authors stringlengths 2 7.06k | published stringlengths 10 10 ⌀ | conference stringlengths 12 47 ⌀ | conference_url_abs stringlengths 16 198 ⌀ | conference_url_pdf stringlengths 27 199 ⌀ | proceeding stringlengths 6 47 ⌀ | taskID stringlengths 7 1.44k | areaID stringclasses 688 values | embedding stringlengths 9.26k 12.5k | umap_embedding stringlengths 29 44 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
03c6393e-dfe0-4a9a-9c0b-f881c617fa33 | a-bayesian-sparse-factor-model-with-adaptive | 2305.18488 | null | https://arxiv.org/abs/2305.18488v1 | https://arxiv.org/pdf/2305.18488v1.pdf | A Bayesian sparse factor model with adaptive posterior concentration | In this paper, we propose a new Bayesian inference method for a high-dimensional sparse factor model that allows both the factor dimensionality and the sparse structure of the loading matrix to be inferred. The novelty is to introduce a certain dependence between the sparsity level and the factor dimensionality, which leads to adaptive posterior concentration while keeping computational tractability. We show that the posterior distribution asymptotically concentrates on the true factor dimensionality, and more importantly, this posterior consistency is adaptive to the sparsity level of the true loading matrix and the noise variance. We also prove that the proposed Bayesian model attains the optimal detection rate of the factor dimensionality in a more general situation than those found in the literature. Moreover, we obtain a near-optimal posterior concentration rate of the covariance matrix. Numerical studies are conducted and show the superiority of the proposed method compared with other competitors. | ['Yongdai Kim', 'Lizhen Lin', 'Ilsang Ohn'] | 2023-05-29 | null | null | null | null | ['bayesian-inference'] | ['methodology'] | [ 1.13319442e-01 -1.37771070e-02 -4.27289277e-01 1.78101942e-01
-4.75074410e-01 -4.93942708e-01 2.06219763e-01 -3.60013634e-01
-9.62064043e-02 5.05237401e-01 2.69852281e-01 -1.96138516e-01
-7.78280973e-01 -3.49141657e-01 -5.28426230e-01 -9.38877463e-01
-1.40260041e-01 3.12474310e-01 -4.08411026e-02 2.54565388e-01
1.34674668e-01 3.30537647e-01 -1.15947926e+00 -5.97963333e-01
9.05270517e-01 1.03862107e+00 2.18582198e-01 4.72472906e-01
4.68821466e-01 7.30060190e-02 -3.64821702e-01 -3.27726215e-01
5.27359843e-01 -3.70518267e-01 -6.00183979e-02 4.28657800e-01
1.02274634e-01 -2.39437968e-01 -2.31448725e-01 1.55943549e+00
2.85470843e-01 2.11642645e-02 7.22365677e-01 -1.01936626e+00
-1.54836446e-01 5.61968088e-01 -1.15509510e+00 3.76643300e-01
1.08925924e-01 -4.02999252e-01 8.96294713e-01 -9.22304153e-01
3.32650632e-01 1.28638482e+00 5.43707013e-01 -1.24169789e-01
-1.16135383e+00 -9.06810760e-01 2.44766235e-01 -1.81073874e-01
-1.82019150e+00 -3.72212470e-01 6.09685063e-01 -5.70739627e-01
8.54395032e-02 -4.87571843e-02 3.58895361e-01 9.27581966e-01
2.68410712e-01 4.92239714e-01 1.06045783e+00 -5.20960927e-01
4.30926830e-01 -2.62845699e-02 3.12663317e-01 8.20331216e-01
9.97095227e-01 1.26335502e-01 -5.03945947e-01 -5.88169217e-01
9.97941673e-01 -9.63168144e-02 -4.52965885e-01 -4.82930720e-01
-9.48062360e-01 8.55252206e-01 -2.45758802e-01 3.21040690e-01
-4.97110724e-01 8.01180452e-02 -4.68106382e-02 -8.79030004e-02
3.11799735e-01 8.29967260e-02 -2.16223806e-01 -1.28308654e-01
-9.81994808e-01 2.54452407e-01 9.56587136e-01 1.09901750e+00
4.24024314e-01 1.32990360e-01 -2.68384904e-01 5.59789419e-01
5.41477859e-01 1.13055599e+00 -3.39403888e-03 -1.02436852e+00
6.87463462e-01 5.82999066e-02 5.52382708e-01 -1.43831348e+00
-2.63059616e-01 -9.31526482e-01 -1.12264311e+00 -4.52109933e-01
6.84666157e-01 -5.25907993e-01 -4.81125593e-01 2.15190697e+00
4.35820818e-01 4.11590815e-01 -1.21941030e-01 9.33941782e-01
-2.17166021e-01 5.47401547e-01 -4.77336168e-01 -7.52419174e-01
1.52270782e+00 -3.88202667e-01 -1.17700207e+00 -2.88114995e-01
1.17092006e-01 -6.80128396e-01 6.74936235e-01 5.80393016e-01
-9.31988239e-01 -2.00972155e-01 -1.18378830e+00 6.94815874e-01
4.60544229e-01 4.46854353e-01 6.26046240e-01 1.07655442e+00
-5.04011691e-01 2.94089541e-02 -9.56581771e-01 -1.93822309e-01
1.18762597e-01 1.99608147e-01 1.58026125e-02 -1.29112512e-01
-9.54737425e-01 2.95065373e-01 4.38964665e-01 2.61485010e-01
-6.05524123e-01 -9.07821536e-01 -4.53785002e-01 3.51366937e-01
5.97589791e-01 -6.57113314e-01 9.35676575e-01 -4.45871443e-01
-1.32488990e+00 1.74410250e-02 -4.27236319e-01 -2.19507292e-01
3.74336928e-01 -3.13148469e-01 -1.38349101e-01 3.75022799e-01
2.51281500e-01 -3.94355774e-01 1.30828893e+00 -1.02505004e+00
-6.58213973e-01 -6.72378242e-01 -1.52287081e-01 -2.74490491e-02
-5.05149007e-01 -2.25799128e-01 -7.30427265e-01 -6.89504623e-01
4.24110144e-01 -9.48391438e-01 -4.32253420e-01 -2.08581895e-01
-4.29999352e-01 5.90148717e-02 3.77641201e-01 -4.96172041e-01
1.52580321e+00 -2.46113062e+00 2.87133873e-01 8.36754560e-01
2.37480521e-01 -2.38901868e-01 1.42328188e-01 3.77313077e-01
6.42790422e-02 -2.92042583e-01 -1.26405984e-01 -3.27446349e-02
8.14506263e-02 -6.85750367e-03 -4.22676086e-01 1.01581323e+00
-1.29485294e-01 1.56348109e-01 -8.16607356e-01 -2.81935334e-01
-1.56722292e-01 3.35434437e-01 -7.00541079e-01 8.09110254e-02
2.97397703e-01 1.85741112e-01 -7.36393094e-01 4.49019492e-01
1.05563748e+00 -6.36294186e-01 8.47270787e-01 -3.30985576e-01
1.64788559e-01 -3.82523447e-01 -1.95542669e+00 1.24301386e+00
-1.67383224e-01 -7.07709091e-03 5.20516872e-01 -9.35232043e-01
8.29696834e-01 2.05606058e-01 4.47763622e-01 -1.53864443e-01
1.81701973e-01 6.61696792e-02 1.00543797e-01 -1.36029512e-01
2.72095412e-01 -3.07209313e-01 -1.45283133e-01 2.69856811e-01
-9.83865745e-03 5.08106172e-01 4.47169453e-01 2.20780298e-01
7.43547082e-01 -3.53766710e-01 6.17468894e-01 -6.57359362e-01
3.09971243e-01 -6.31697118e-01 8.29769611e-01 1.06358445e+00
4.88369688e-02 -2.44956072e-02 8.92785192e-01 3.26396316e-01
-1.05725503e+00 -1.16498971e+00 -5.47658741e-01 6.50852382e-01
2.21410275e-01 -5.46936393e-01 -7.22942233e-01 -2.83358067e-01
1.03365920e-01 3.37180078e-01 -6.48692787e-01 1.24248136e-02
-2.66811967e-01 -1.09425080e+00 3.20257902e-01 2.99893439e-01
2.95388728e-01 3.45445007e-01 -1.53533101e-01 3.29395235e-02
-7.03938231e-02 -1.39964604e+00 -4.80586201e-01 -5.30075803e-02
-9.15880322e-01 -1.10885572e+00 -6.75534666e-01 -2.68948734e-01
7.92122066e-01 3.52603316e-01 3.79499674e-01 -4.26851422e-01
3.34560633e-01 3.98848683e-01 -1.89749628e-01 -1.62564605e-01
-1.10648461e-01 5.00952341e-02 3.84803146e-01 4.77400452e-01
5.17407656e-02 -6.27398014e-01 -4.85094965e-01 5.27920961e-01
-8.61854196e-01 -1.51015505e-01 5.51095843e-01 8.20887625e-01
4.41188991e-01 6.80627108e-01 5.50073981e-01 -7.37475991e-01
6.86331868e-01 -7.02895403e-01 -1.15904093e+00 4.66188937e-02
-7.77999341e-01 1.80305213e-01 5.37904859e-01 -4.26769197e-01
-1.18956482e+00 1.62683390e-02 3.09107840e-01 -4.42171991e-01
1.98881835e-01 7.11349666e-01 -3.15997750e-01 2.12600589e-01
1.66960269e-01 -1.92887932e-02 -7.06637874e-02 -6.09548151e-01
3.54683191e-01 4.77684200e-01 4.74551052e-01 -8.10327053e-01
9.96012807e-01 6.43223941e-01 4.88584101e-01 -8.56396973e-01
-1.05446768e+00 -3.46808940e-01 -3.75366986e-01 1.30360099e-02
2.42650211e-01 -1.03584278e+00 -6.99024558e-01 1.43076554e-01
-7.48396575e-01 2.78297246e-01 1.66563448e-02 1.02310050e+00
-5.13036966e-01 8.23032022e-01 -5.12011588e-01 -1.25046539e+00
6.97111189e-02 -9.09825981e-01 8.26222181e-01 -4.07685246e-03
6.62972331e-02 -8.71858299e-01 1.53652430e-01 5.62798530e-02
5.25361821e-02 5.89744598e-02 7.89132774e-01 -3.21254313e-01
-3.50085467e-01 -3.15778792e-01 -4.10821021e-01 -3.97420898e-02
6.48074523e-02 -3.21404070e-01 -3.20272684e-01 -5.99515498e-01
3.36614370e-01 4.39788908e-01 6.34414375e-01 7.52852559e-01
8.50150049e-01 -5.24390876e-01 -5.03436208e-01 5.14800966e-01
1.37645364e+00 7.86957592e-02 3.22320491e-01 -6.37233481e-02
5.37722766e-01 4.05021667e-01 6.69377625e-01 1.20100343e+00
-8.20872337e-02 7.91398168e-01 -1.05782235e-02 3.70500833e-01
3.62046957e-01 -4.04913038e-01 1.85252458e-01 1.15535867e+00
7.39392117e-02 -2.67546326e-01 -5.59897125e-01 4.29601669e-01
-2.01194763e+00 -8.05743694e-01 -6.92623258e-02 2.57130384e+00
5.47824740e-01 3.30069248e-04 1.76311225e-01 3.13703686e-01
8.99954677e-01 -9.93971825e-02 -6.34402573e-01 1.06057964e-01
-2.57552952e-01 -2.04387829e-02 8.79167080e-01 5.48590243e-01
-8.83461356e-01 4.56742465e-01 7.52075863e+00 1.02046168e+00
-3.68380427e-01 1.63334653e-01 2.50478178e-01 -9.07752216e-02
-1.78463414e-01 2.75281165e-02 -1.27526891e+00 7.01107919e-01
9.05020416e-01 -4.08510894e-01 3.66966486e-01 7.88002193e-01
4.42792177e-01 -6.00848868e-02 -7.18596995e-01 1.03182662e+00
3.09743192e-02 -8.21036220e-01 -1.00929171e-01 5.81311166e-01
7.29786038e-01 -4.97337669e-01 2.45034415e-02 -1.96028024e-01
1.64301559e-01 -5.25794446e-01 5.08363843e-01 6.16743267e-01
7.05075741e-01 -1.16180491e+00 7.51084685e-01 3.63327205e-01
-1.12595487e+00 -5.94497800e-01 -5.78519881e-01 -1.18009314e-01
3.61942947e-01 1.34451461e+00 -5.41122615e-01 5.00346959e-01
4.70007837e-01 7.04120517e-01 -2.04186872e-01 1.01108456e+00
-2.34427616e-01 8.02506030e-01 -7.06798375e-01 1.65098101e-01
2.80831140e-02 -6.11200213e-01 8.33743691e-01 1.08708704e+00
6.94533944e-01 4.84785959e-02 3.12435120e-01 6.37706816e-01
6.91960081e-02 1.24923036e-01 -3.71899545e-01 -1.57491267e-01
8.66494834e-01 9.83793914e-01 -7.09888935e-01 -2.18435243e-01
-3.33056509e-01 4.58217174e-01 1.13756433e-01 6.12569749e-01
-5.75320542e-01 -2.59124011e-01 6.88400388e-01 4.07095253e-02
9.75249052e-01 -3.21420074e-01 -9.17791501e-02 -1.24992263e+00
1.88816980e-01 -9.64731872e-01 5.12294114e-01 -1.51476264e-01
-1.41299987e+00 3.88769656e-02 3.42389345e-01 -1.02924097e+00
-4.06334341e-01 -6.07100070e-01 5.78582510e-02 6.66297257e-01
-9.63328600e-01 -4.63192254e-01 2.34579116e-01 5.56356728e-01
-2.02239528e-02 -3.04156125e-01 3.57730180e-01 3.41822267e-01
-8.85274768e-01 6.54889882e-01 7.12853312e-01 -8.33127499e-02
2.53876746e-01 -8.76750648e-01 -3.21869224e-01 1.09714830e+00
-2.35946149e-01 9.51398492e-01 1.03515291e+00 -6.80857003e-01
-1.80688155e+00 -5.47159851e-01 3.08008134e-01 6.16953038e-02
1.16817021e+00 -4.03079629e-01 -5.01003563e-01 5.52162290e-01
-1.43000215e-01 -3.37660521e-01 9.61900890e-01 4.84539390e-01
-4.71092850e-01 -1.51329160e-01 -9.86544490e-01 5.59273958e-01
8.56450379e-01 -2.75357872e-01 -3.51987183e-01 2.61487782e-01
5.30667067e-01 -2.36238256e-01 -1.08976305e+00 4.37610358e-01
8.60547006e-01 -6.95996821e-01 8.77683520e-01 -5.29612422e-01
-3.26513536e-02 -5.65390050e-01 -5.92623293e-01 -9.19390619e-01
-7.57053733e-01 -7.46855974e-01 -6.46977901e-01 1.11676729e+00
1.22377008e-01 -6.03317797e-01 6.40003681e-01 2.75440127e-01
4.02605772e-01 -5.40935576e-01 -1.22626626e+00 -9.92484212e-01
-2.03419641e-01 -4.02043104e-01 2.70219207e-01 7.65478313e-01
2.26016007e-02 5.08737147e-01 -9.54888582e-01 7.70463824e-01
1.21686888e+00 2.19900355e-01 6.07715905e-01 -1.30102813e+00
-9.32652771e-01 -1.05602778e-01 -3.00430000e-01 -1.46592879e+00
1.71702340e-01 -5.40267944e-01 -2.59181827e-01 -8.37207437e-01
5.90283155e-01 -2.03303978e-01 -3.65098625e-01 -1.69304594e-01
-2.21311435e-01 -9.39859450e-02 1.01866975e-01 3.03810745e-01
-4.82677996e-01 5.57740688e-01 1.28333402e+00 2.56922901e-01
-1.26992196e-01 1.68345273e-01 -1.04455435e+00 7.36758828e-01
4.67426032e-01 -5.92839301e-01 -5.88141561e-01 -3.54670845e-02
6.21219814e-01 4.16827619e-01 4.04357398e-03 -7.96614587e-01
3.66434269e-02 -1.01323761e-01 3.49028736e-01 -6.84518456e-01
2.03673914e-01 -8.49797547e-01 2.40135580e-01 4.73510414e-01
-1.38999209e-01 -2.31202990e-01 -7.04068094e-02 1.35224771e+00
2.46756077e-01 -4.22159344e-01 7.23890603e-01 4.38248545e-01
2.15925992e-01 2.16806278e-01 -5.67308903e-01 2.86896173e-02
6.66633785e-01 1.24785356e-01 -7.11484551e-02 -5.94589651e-01
-7.00827599e-01 3.81154679e-02 4.51482870e-02 1.40201911e-01
2.82020897e-01 -1.44637215e+00 -6.24372005e-01 2.15620220e-01
-9.26068127e-02 -6.75356746e-01 2.88200825e-01 1.13530517e+00
2.44805425e-01 5.22174716e-01 2.07141832e-01 -7.10666716e-01
-8.16543579e-01 6.67784214e-01 -6.27488717e-02 -2.82885045e-01
-5.37579119e-01 2.31949940e-01 3.43351990e-01 4.05528443e-03
2.23018959e-01 -7.40826949e-02 4.81245518e-02 -5.27960062e-02
9.60636854e-01 5.74452102e-01 -3.45284402e-01 -4.39466119e-01
-2.70695537e-01 6.16266310e-01 2.53749453e-02 -1.21792465e-01
1.04468131e+00 -5.19293725e-01 -1.73574418e-01 4.10600573e-01
9.57377136e-01 3.92852098e-01 -1.25521982e+00 -4.97712284e-01
-2.97247134e-02 -9.66133952e-01 2.69359529e-01 -1.66666150e-01
-1.12177730e+00 2.56275862e-01 7.15505242e-01 1.62745655e-01
9.40110683e-01 6.42205402e-03 2.86239266e-01 4.11343515e-01
5.27816653e-01 -9.27132189e-01 -1.88668802e-01 3.10609788e-01
4.18114066e-01 -6.35058284e-01 3.92276257e-01 -6.78987265e-01
-1.14787564e-01 8.24508369e-01 8.50829631e-02 -1.83954090e-01
9.46393609e-01 4.09898430e-01 -6.43696308e-01 -5.29154465e-02
-5.69323242e-01 -5.28000891e-02 2.39469483e-01 6.01330698e-01
1.72738716e-01 1.95571065e-01 -6.98755741e-01 1.06572938e+00
-1.60250738e-01 -9.59875733e-02 5.27492464e-01 3.51439655e-01
-4.65613037e-01 -9.23387170e-01 -7.54379511e-01 4.56621408e-01
-7.05383956e-01 -7.60522345e-03 4.15354557e-02 5.10404766e-01
-3.15068156e-01 1.10931993e+00 -5.64533994e-02 1.00933850e-01
1.55974090e-01 -2.94513226e-01 5.53512037e-01 -2.59980738e-01
3.97198945e-01 7.51502693e-01 2.66223010e-02 -4.25032765e-01
-3.76428455e-01 -6.87912583e-01 -6.30947828e-01 -3.57720792e-01
-5.10031998e-01 5.65861046e-01 4.69963402e-01 1.04847431e+00
4.35218960e-01 1.16787300e-01 7.92973936e-01 -3.76494586e-01
-7.93350220e-01 -8.68642092e-01 -1.30107415e+00 -5.50129078e-02
1.75780207e-01 -1.00840747e+00 -7.54621089e-01 -2.81068474e-01] | [7.046165943145752, 4.5426459312438965] |
5e0a437f-29a6-4ae2-9bf2-11bc499b8065 | towards-building-text-to-speech-systems-for | 2211.09536 | null | https://arxiv.org/abs/2211.09536v3 | https://arxiv.org/pdf/2211.09536v3.pdf | Towards Building Text-To-Speech Systems for the Next Billion Users | Deep learning based text-to-speech (TTS) systems have been evolving rapidly with advances in model architectures, training methodologies, and generalization across speakers and languages. However, these advances have not been thoroughly investigated for Indian language speech synthesis. Such investigation is computationally expensive given the number and diversity of Indian languages, relatively lower resource availability, and the diverse set of advances in neural TTS that remain untested. In this paper, we evaluate the choice of acoustic models, vocoders, supplementary loss functions, training schedules, and speaker and language diversity for Dravidian and Indo-Aryan languages. Based on this, we identify monolingual models with FastPitch and HiFi-GAN V1, trained jointly on male and female speakers to perform the best. With this setup, we train and evaluate TTS models for 13 languages and find our models to significantly improve upon existing models in all languages as measured by mean opinion scores. We open-source all models on the Bhashini platform. | ['Karthik Nandakumar', 'Mitesh M. Khapra', 'Pratyush Kumar', 'Praveen S V', 'Gokul Karthik Kumar'] | 2022-11-17 | null | null | null | null | ['speech-synthesis-bodo', 'speech-synthesis-rajasthani', 'speech-synthesis-marathi', 'speech-synthesis-kannada', 'text-to-speech-synthesis', 'speech-synthesis-odia', 'speech-synthesis-manipuri', 'speech-synthesis-bengali', 'speech-synthesis-assamese', 'speech-synthesis-hindi', 'speech-synthesis-gujarati', 'speech-synthesis-tamil', 'speech-synthesis-malayalam', 'speech-synthesis-telugu'] | ['speech', 'speech', 'speech', 'speech', 'speech', 'speech', 'speech', 'speech', 'speech', 'speech', 'speech', 'speech', 'speech', 'speech'] | [-2.48201385e-01 -1.13367334e-01 -1.16168279e-02 -5.22112370e-01
-1.31636751e+00 -7.06795514e-01 7.21595228e-01 -4.27264273e-01
-1.91660240e-01 6.16998672e-01 4.54056323e-01 -7.59070635e-01
3.73823106e-01 -6.63356185e-02 -5.02693951e-01 -5.38392067e-01
1.58394054e-01 7.08839536e-01 -1.40325740e-01 -4.04819131e-01
-3.16332579e-01 3.80849123e-01 -1.09703338e+00 2.44270377e-02
8.92424881e-01 7.24035680e-01 2.16022983e-01 8.55621517e-01
5.58628961e-02 6.83756590e-01 -9.06618774e-01 -5.85731983e-01
7.25331083e-02 -7.12946415e-01 -4.75451440e-01 -1.79915085e-01
5.27529240e-01 -2.21021846e-01 -4.15564328e-01 5.14254808e-01
1.07780659e+00 8.00132155e-02 6.49422169e-01 -7.20346987e-01
-9.13877845e-01 1.04586923e+00 -2.83928663e-01 1.69443473e-01
-1.04344673e-01 1.72327518e-01 1.00506043e+00 -1.06970358e+00
4.13451701e-01 1.58473969e+00 4.76491153e-01 1.00106776e+00
-1.25229394e+00 -7.98940718e-01 9.03686974e-03 9.13531855e-02
-1.32436502e+00 -1.11778855e+00 7.16305614e-01 -1.86072141e-01
1.11673439e+00 2.91631460e-01 3.64615947e-01 1.58700001e+00
-4.47820760e-02 9.29119706e-01 1.03534937e+00 -6.09725118e-01
1.85958579e-01 4.51428354e-01 -4.90401328e-01 4.59258318e-01
-2.57127762e-01 4.66231890e-02 -7.66716242e-01 1.77589685e-01
4.14215386e-01 -7.73477912e-01 1.64168049e-02 1.54717639e-01
-8.99290085e-01 9.38758731e-01 -2.23549809e-02 3.81323218e-01
-2.94447653e-02 1.90012399e-02 4.37201500e-01 4.97422576e-01
6.25372946e-01 3.80104482e-01 -6.23453617e-01 -1.62929684e-01
-1.18714011e+00 1.67633355e-01 8.56719315e-01 9.65155363e-01
2.31690690e-01 1.09689319e+00 -9.33353379e-02 1.59401584e+00
3.37774843e-01 9.76691782e-01 4.90259767e-01 -7.50898719e-01
5.04927337e-01 -1.88953251e-01 -4.24289435e-01 -3.92250389e-01
-8.73530731e-02 -6.42602026e-01 -6.81323886e-01 3.36803086e-02
1.53375402e-01 -5.00421405e-01 -1.21971667e+00 1.80460393e+00
-3.24509777e-02 -1.78010553e-01 2.87419647e-01 5.75114787e-01
8.49981546e-01 9.59879935e-01 7.52249360e-02 -8.63095373e-02
9.73552167e-01 -1.03685069e+00 -6.89122558e-01 -4.13502514e-01
5.60577333e-01 -1.15014458e+00 1.26995718e+00 3.69664758e-01
-1.30123711e+00 -6.78347826e-01 -1.04259717e+00 -2.20585078e-01
-2.15607703e-01 4.99156475e-01 2.80717105e-01 1.14303100e+00
-1.28784275e+00 6.04950860e-02 -8.51953506e-01 -4.13163126e-01
1.06692620e-01 3.86403322e-01 -8.63269717e-02 7.03115165e-02
-1.16471541e+00 1.05691171e+00 1.34778813e-01 1.30425081e-01
-1.16074300e+00 -4.89349633e-01 -6.57951891e-01 -1.78128686e-02
-1.29897110e-02 -3.47142607e-01 1.54339981e+00 -9.15762663e-01
-2.04744816e+00 5.63102484e-01 -2.34145373e-01 -5.27274013e-01
3.47132236e-01 4.27640006e-02 -7.74405420e-01 -3.52490723e-01
-2.44860888e-01 9.21081245e-01 6.27775073e-01 -1.26484573e+00
-5.35033882e-01 -1.70747027e-01 -5.44931471e-01 3.64159375e-01
-4.33628112e-01 4.28259760e-01 -4.72614169e-01 -7.77289391e-01
-2.48863801e-01 -1.04685378e+00 -8.43537748e-02 -5.81181645e-01
-3.48619759e-01 -1.44147471e-01 7.93091297e-01 -1.06783164e+00
1.07177830e+00 -2.36295652e+00 1.51877046e-01 1.19770989e-01
-4.65203792e-01 3.05726498e-01 -2.93508172e-01 3.65198284e-01
3.49985957e-01 2.75323093e-01 -1.49580985e-01 -7.50561476e-01
1.44185022e-01 3.86183590e-01 -3.92159492e-01 7.87985101e-02
3.19769114e-01 6.99189305e-01 -3.31653357e-01 -4.34988886e-01
2.60275960e-01 8.17851543e-01 -5.61406314e-01 2.83812732e-01
-2.60082573e-01 8.20981383e-01 -1.65278092e-01 8.45181346e-01
3.35782290e-01 4.64978814e-01 1.88087448e-01 2.85214782e-01
-2.08143920e-01 8.06837022e-01 -7.99245656e-01 1.60484576e+00
-7.68970072e-01 7.35324562e-01 3.03513020e-01 -7.90880322e-01
1.18614614e+00 6.76837325e-01 6.95807710e-02 -8.86712313e-01
8.69516209e-02 8.01047027e-01 4.64640737e-01 -2.18470529e-01
4.76765335e-01 -1.91013962e-01 1.43759302e-03 2.60640844e-03
4.46100771e-01 -5.70905864e-01 1.11937478e-01 -1.68268830e-01
6.32648647e-01 -4.63171490e-02 -2.29580358e-01 -3.48830789e-01
2.51974136e-01 -2.74552971e-01 2.89783925e-01 7.46141613e-01
-3.16886790e-02 7.07438529e-01 1.75003350e-01 -1.55725852e-02
-1.30017471e+00 -1.21950436e+00 -1.74685925e-01 1.53929722e+00
-6.80298924e-01 -2.32544437e-01 -8.38384330e-01 -1.98657826e-01
-3.24136972e-01 1.05671597e+00 -1.20858490e-01 1.63838789e-01
-8.87645483e-01 -5.65354645e-01 9.12079692e-01 4.12867010e-01
8.81238133e-02 -1.25729072e+00 2.46681944e-01 4.41569448e-01
-6.62627965e-02 -1.22882426e+00 -5.69167852e-01 3.06978881e-01
-5.86351156e-01 -1.97364166e-01 -1.04338002e+00 -1.10366309e+00
6.09839074e-02 -3.26181650e-01 1.09745741e+00 -4.95989352e-01
1.27165109e-01 1.50350019e-01 -1.54853255e-01 -7.64642894e-01
-9.55082834e-01 5.56027889e-01 2.10431665e-01 -1.16415918e-01
-2.95874216e-02 -3.70887905e-01 -1.27532423e-01 1.18636280e-01
-5.31161964e-01 -2.03567147e-01 6.12187326e-01 6.44651949e-01
4.12953883e-01 -1.47398204e-01 7.01028526e-01 -6.95929468e-01
6.00550771e-01 -3.93150002e-01 -4.96062398e-01 2.74630815e-01
-4.35596406e-01 1.62181426e-02 6.60345793e-01 -2.93076038e-01
-1.23348534e+00 -2.63008088e-01 -8.21207225e-01 -1.50697157e-01
-6.38117343e-02 5.51932335e-01 -2.60488212e-01 2.03795716e-01
6.53973639e-01 2.42783412e-01 -3.60667855e-01 -5.58623075e-01
3.60667497e-01 1.10309458e+00 5.73482633e-01 -6.89834058e-01
5.63066125e-01 -1.24781810e-01 -4.28766638e-01 -1.33976150e+00
-5.40940285e-01 5.44905178e-02 -3.66892308e-01 -5.21981083e-02
7.33779073e-01 -1.22114611e+00 -3.33392411e-01 5.54272413e-01
-1.08482921e+00 -4.83062744e-01 -9.87200160e-03 6.55920446e-01
-3.23347270e-01 -1.28502071e-01 -6.72977269e-01 -1.05499923e+00
-5.01895964e-01 -1.51009142e+00 1.04696691e+00 4.50500511e-02
-2.44190231e-01 -1.04028440e+00 1.21829458e-01 4.18309391e-01
8.69307935e-01 -1.45466223e-01 1.07734954e+00 -7.42449224e-01
-3.77735376e-01 1.30503237e-01 2.95505762e-01 7.36037850e-01
6.03432357e-02 2.52895802e-01 -1.12382376e+00 -4.04751003e-01
-1.91688046e-01 -4.73620027e-01 4.80096310e-01 6.52248085e-01
8.41926336e-01 -2.41085708e-01 1.13332026e-01 7.09501386e-01
7.26562142e-01 4.84988183e-01 3.27802002e-01 2.40917914e-02
7.69415140e-01 7.38657057e-01 -1.02451257e-01 4.90156077e-02
4.24636573e-01 6.81042910e-01 -5.66531830e-02 -2.25145832e-01
-4.31104749e-01 -2.12203786e-01 7.44478285e-01 1.67273009e+00
1.40074670e-01 -6.35077178e-01 -1.01288593e+00 6.76262379e-01
-1.16968679e+00 -6.34803236e-01 2.04965457e-01 2.00986147e+00
1.08416367e+00 1.95431247e-01 2.86947906e-01 -2.15510935e-01
5.37207901e-01 2.74482429e-01 -4.29055274e-01 -9.45853531e-01
-6.28869772e-01 5.37826180e-01 3.13831717e-01 6.11857474e-01
-7.00469673e-01 1.34094393e+00 6.79826403e+00 9.53264594e-01
-1.57635045e+00 2.25264683e-01 9.51932609e-01 -2.28714481e-01
-5.86687863e-01 -1.69531748e-01 -1.08804762e+00 3.14049333e-01
1.63278306e+00 1.34871155e-01 7.65806377e-01 7.03905106e-01
2.11967856e-01 5.91815591e-01 -1.04973114e+00 7.36277342e-01
1.84764311e-01 -9.85286415e-01 -6.02449942e-03 -9.16262195e-02
9.00278032e-01 5.99232495e-01 4.11894590e-01 7.20320523e-01
5.00182629e-01 -1.25747919e+00 1.10884523e+00 -8.80268663e-02
9.29615200e-01 -8.22689414e-01 4.72336411e-01 -5.18647097e-02
-9.33011532e-01 5.09268008e-02 -1.16332829e-01 3.64429951e-01
1.43099278e-01 1.02522299e-01 -1.03623116e+00 2.93935478e-01
6.15624726e-01 1.71963364e-01 -4.38042551e-01 5.76190114e-01
-1.78064391e-01 1.53170204e+00 -5.10571122e-01 -1.56296074e-01
4.06325340e-01 2.18942054e-02 3.85772318e-01 1.41881132e+00
5.72635531e-01 -3.83090138e-01 1.69533432e-01 5.82823098e-01
-1.52577132e-01 4.26235229e-01 -4.28116024e-01 -2.70098746e-01
6.59883261e-01 8.55796933e-01 -3.04659486e-01 -1.44891709e-01
-5.18823981e-01 7.31672168e-01 2.38159776e-01 6.18403852e-01
-6.63850665e-01 -1.67080641e-01 5.98336160e-01 -5.43953739e-02
2.49460876e-01 -6.45898223e-01 -2.52034098e-01 -8.88617277e-01
-1.84916575e-02 -1.30003047e+00 1.10521458e-01 -5.59939384e-01
-1.19496369e+00 1.09325588e+00 -1.43498704e-01 -6.33202016e-01
-5.99060059e-01 -5.94902933e-01 -5.68103254e-01 1.17342925e+00
-1.39867234e+00 -1.46566963e+00 4.34080452e-01 4.03722405e-01
9.24244225e-01 -6.97934926e-01 8.73147905e-01 5.20769477e-01
-6.01065695e-01 1.03394997e+00 3.66861910e-01 3.12928334e-02
7.48592675e-01 -1.08427715e+00 9.10817742e-01 7.89469123e-01
3.94238114e-01 4.04734910e-01 7.00484693e-01 -3.07321668e-01
-1.19403815e+00 -9.72029090e-01 1.08891916e+00 -3.51682514e-01
6.09241724e-01 -9.34785485e-01 -8.19894552e-01 8.66306543e-01
6.02219105e-01 -3.99154216e-01 6.54084623e-01 4.17123497e-01
-2.92234898e-01 -2.70235300e-01 -6.79798007e-01 9.32422101e-01
6.81544542e-01 -7.34747946e-01 -8.14841315e-02 1.99220970e-01
7.71803856e-01 -4.42259252e-01 -7.26484656e-01 2.78224319e-01
5.43597460e-01 -6.87757850e-01 7.37542272e-01 -3.69590223e-01
5.23266867e-02 2.48739701e-02 -5.94435275e-01 -1.59760177e+00
-1.22833930e-01 -8.89267445e-01 3.55881900e-01 1.67317176e+00
9.90093052e-01 -5.42159319e-01 5.13623416e-01 2.04842776e-01
-6.12609386e-01 -6.33131564e-01 -1.07478452e+00 -8.11505556e-01
5.32676816e-01 -4.05768335e-01 5.18596351e-01 7.91240215e-01
-4.43269104e-01 6.81996524e-01 -6.83736861e-01 4.77285273e-02
2.76392728e-01 -3.73015404e-01 6.82583451e-01 -7.94917881e-01
-5.00602782e-01 -6.01160884e-01 2.63584219e-02 -8.15890253e-01
3.94229978e-01 -9.65027809e-01 1.63914382e-01 -1.37379491e+00
-2.32807964e-01 -5.58707237e-01 -1.89899653e-01 5.74772477e-01
-9.09036696e-02 1.58657625e-01 2.03122854e-01 8.95518064e-02
4.64144684e-02 7.64046192e-01 8.39949906e-01 -2.02575207e-01
-3.42977375e-01 -4.62875590e-02 -5.39137065e-01 4.70303655e-01
9.61401105e-01 -2.71587282e-01 -4.58762020e-01 -1.01468968e+00
-2.89877743e-01 1.61858737e-01 -2.88805753e-01 -1.05281186e+00
4.37349407e-03 2.78436788e-03 3.77599001e-01 -3.70345920e-01
6.35892987e-01 -4.67177451e-01 1.30561158e-01 2.08021462e-01
-5.17935634e-01 2.64775395e-01 4.27133262e-01 -1.16370857e-01
-2.51868367e-01 1.66495498e-02 9.29765522e-01 8.04307908e-02
-4.19089109e-01 2.37536743e-01 -5.37496924e-01 2.83458382e-01
3.50064814e-01 -1.31616024e-02 -4.14901376e-02 -6.16666079e-01
-6.03114665e-01 1.41977305e-02 5.46213193e-03 7.31749833e-01
2.44643241e-01 -1.25302756e+00 -1.37595057e+00 4.07712340e-01
5.10833822e-02 -3.62149060e-01 1.25109881e-01 4.38077271e-01
-4.56045002e-01 5.69660306e-01 -1.00086585e-01 -5.93234658e-01
-1.06716573e+00 4.17387262e-02 2.63693869e-01 4.71898690e-02
-1.40661567e-01 1.06274986e+00 6.59026131e-02 -9.33413863e-01
4.52332139e-01 -1.50067627e-01 1.29500806e-01 -1.91230580e-01
-7.84293711e-02 1.97263747e-01 2.24089414e-01 -9.17468011e-01
-2.90286779e-01 3.18933785e-01 -2.11014166e-01 -5.93560159e-01
1.09386790e+00 -1.41720727e-01 3.30050439e-01 7.11779833e-01
1.10931468e+00 3.41700703e-01 -9.70377326e-01 -7.38718212e-02
-7.30972961e-02 1.59925241e-02 5.76149151e-02 -1.00930643e+00
-1.07754827e+00 1.11074007e+00 5.09717107e-01 2.39277221e-02
8.62556338e-01 6.39075413e-02 8.48565459e-01 2.62883753e-01
-2.46227849e-02 -1.16632044e+00 -2.68466860e-01 8.74872804e-01
9.24455941e-01 -1.16157520e+00 -4.37105387e-01 5.50876409e-02
-1.01188993e+00 7.52689600e-01 5.64140081e-01 3.54052156e-01
5.78852355e-01 5.26118279e-01 6.03273571e-01 3.64415109e-01
-8.89883459e-01 5.88993717e-05 2.55001873e-01 4.96846080e-01
9.88431156e-01 3.72485906e-01 -1.82526857e-02 1.84811920e-01
-9.13401186e-01 -6.25560522e-01 2.27300867e-01 4.92623001e-01
-2.57619649e-01 -1.26934123e+00 -2.67112732e-01 2.01877356e-01
-7.48364925e-01 -3.75247061e-01 -3.61788005e-01 6.57848656e-01
-8.60582292e-02 1.20080769e+00 -2.96962485e-02 -3.01270902e-01
3.70676190e-01 2.46109053e-01 4.37097818e-01 -5.82094967e-01
-5.97633779e-01 5.50125659e-01 4.52102065e-01 2.16427296e-01
3.38893607e-02 -7.56247401e-01 -1.03017771e+00 -2.78942972e-01
-3.67649704e-01 1.86311051e-01 1.24646449e+00 8.48996699e-01
2.69973606e-01 5.69432735e-01 7.35773861e-01 -6.96992040e-01
-6.50504887e-01 -1.31924284e+00 -5.30819535e-01 -1.36451647e-01
2.88842946e-01 -1.19710572e-01 -2.64868140e-01 3.18041071e-02] | [14.645650863647461, 6.852306365966797] |
697b939c-430d-4e59-b101-71c4d8ec7c77 | a-novel-strategy-for-improving-robustness-in | 2305.09407 | null | https://arxiv.org/abs/2305.09407v1 | https://arxiv.org/pdf/2305.09407v1.pdf | A Novel Strategy for Improving Robustness in Computer Vision Manufacturing Defect Detection | Visual quality inspection in high performance manufacturing can benefit from automation, due to cost savings and improved rigor. Deep learning techniques are the current state of the art for generic computer vision tasks like classification and object detection. Manufacturing data can pose a challenge for deep learning because data is highly repetitive and there are few images of defects or deviations to learn from. Deep learning models trained with such data can be fragile and sensitive to context, and can under-detect new defects not found in the training data. In this work, we explore training defect detection models to learn specific defects out of context, so that they are more likely to be detected in new situations. We demonstrate how models trained on diverse images containing a common defect type can pick defects out in new circumstances. Such generic models could be more robust to new defects not found data collected for training, and can reduce data collection impediments to implementing visual inspection on production lines. Additionally, we demonstrate that object detection models trained to predict a label and bounding box outperform classifiers that predict a label only on held out test data typical of manufacturing inspection tasks. Finally, we studied the factors that affect generalization in order to train models that work under a wider range of conditions. | ['Andrew E. Marble', 'Ahmad Mohamad Mezher'] | 2023-05-16 | null | null | null | null | ['defect-detection'] | ['computer-vision'] | [ 5.00842571e-01 -6.86482489e-02 3.24334681e-01 -5.28983712e-01
-4.65348870e-01 -2.71701068e-01 -2.16514412e-02 5.36137223e-01
-2.42235959e-02 2.89898247e-01 -4.88382488e-01 -9.96155292e-02
-1.80361480e-01 -8.24540019e-01 -1.06230664e+00 -3.84309500e-01
-2.45143995e-01 5.38419008e-01 2.21334398e-01 -2.48604134e-01
3.15148950e-01 6.90735102e-01 -1.94859588e+00 7.56437957e-01
7.69346774e-01 9.75714326e-01 7.41112351e-01 6.16310060e-01
3.42073262e-01 5.14906228e-01 -1.23045337e+00 1.56380445e-01
3.79981846e-01 -7.05555454e-02 -5.80746591e-01 6.58314407e-01
7.93833375e-01 -6.01419032e-01 2.65143663e-02 8.92219126e-01
2.75523007e-01 -7.31661022e-02 5.70404947e-01 -1.02605855e+00
-9.58311498e-01 -1.26214504e-01 -4.84317183e-01 2.88199306e-01
1.76681995e-01 6.46893501e-01 7.02841759e-01 -8.27402890e-01
5.03863096e-01 1.26937127e+00 8.64038765e-01 5.12210429e-01
-1.14665747e+00 -3.41924012e-01 1.21285982e-01 1.06640920e-01
-5.97666264e-01 -1.83020219e-01 7.43695676e-01 -6.20751500e-01
1.02682745e+00 -2.66222972e-02 6.82626963e-01 9.14736390e-01
5.59538662e-01 7.41211176e-01 8.51347923e-01 -4.95610148e-01
2.00915486e-01 1.47445248e-02 3.84632982e-02 1.02386034e+00
5.62613070e-01 4.26210761e-01 -1.69411078e-01 2.95595288e-01
9.99176681e-01 4.01485980e-01 -1.39373705e-01 -4.86367792e-01
-9.52220082e-01 9.36634898e-01 6.38930857e-01 2.50115305e-01
-3.40838730e-01 -1.73492581e-02 6.23959661e-01 7.65879333e-01
3.83085459e-01 1.24031246e+00 -8.78521442e-01 1.51793808e-01
-6.55631363e-01 1.11311391e-01 5.71666777e-01 7.77091682e-01
7.72129178e-01 2.60795832e-01 5.23548536e-02 1.05158377e+00
8.42293873e-02 2.32863531e-01 1.32802561e-01 -9.21464384e-01
1.23797953e-01 6.58050776e-01 -1.60253253e-02 -8.98696482e-01
-5.52405953e-01 -4.60619539e-01 -3.43996257e-01 8.20664167e-01
3.15928161e-01 -1.09668210e-01 -1.45616162e+00 9.86341298e-01
3.95589545e-02 -5.32731712e-01 -1.13654412e-01 7.07864106e-01
5.17632008e-01 4.67801452e-01 -4.01504993e-01 1.12703048e-01
1.04065740e+00 -7.60167062e-01 -5.57345688e-01 -8.31324399e-01
8.96794260e-01 -8.64602685e-01 1.21707773e+00 9.50568616e-01
-8.74836385e-01 -1.15292358e+00 -1.43041766e+00 2.14584887e-01
-5.14122188e-01 3.54260266e-01 7.52784312e-01 3.55464816e-01
-8.63007367e-01 1.00953114e+00 -7.33380973e-01 -4.03570265e-01
7.42771268e-01 3.97518963e-01 -3.46815825e-01 -7.33981669e-01
-5.55830181e-01 9.94762123e-01 4.11706448e-01 3.70934159e-01
-1.41744673e+00 -6.51167750e-01 -1.10464156e+00 -1.56010702e-01
4.69060659e-01 -2.21672505e-01 1.25300097e+00 -1.02992070e+00
-7.72995830e-01 8.31782877e-01 4.24733639e-01 -2.93923229e-01
1.74452856e-01 -4.22388583e-01 -5.08912504e-01 1.00034207e-01
1.43327519e-01 6.38891578e-01 1.10691440e+00 -1.80199230e+00
-8.42250109e-01 -4.50364888e-01 1.66469976e-01 -4.24042374e-01
-8.24371502e-02 -1.46475464e-01 8.40143710e-02 -2.30378494e-01
1.98303774e-01 -6.32140338e-01 -1.79302007e-01 1.28704563e-01
-2.70665407e-01 -1.53014511e-01 1.27933204e+00 -8.17486584e-01
4.95131195e-01 -2.06590009e+00 -3.50740105e-01 -4.05458957e-02
9.53849033e-02 5.93616009e-01 -1.50248364e-01 4.29032207e-01
-7.31959119e-02 5.66579867e-03 9.33788493e-02 -3.24678421e-03
-2.71735311e-01 3.94230187e-01 1.01955667e-01 3.60378116e-01
9.33060646e-01 4.12069172e-01 -9.84221995e-01 2.02698018e-02
4.62080657e-01 -5.37587032e-02 -5.44349790e-01 2.79188484e-01
-4.42394078e-01 4.64645058e-01 1.13191321e-01 1.20582521e+00
5.49661219e-01 -1.27866641e-01 -1.87908769e-01 -3.51925462e-01
1.64808124e-01 -1.49112910e-01 -9.23057199e-01 1.41462362e+00
-8.21361899e-01 7.82435477e-01 1.06460273e-01 -1.30089283e+00
1.45580876e+00 6.10235296e-02 1.75591350e-01 -6.72057092e-01
5.02962656e-02 2.14650616e-01 4.34608817e-01 -1.09910417e+00
3.69546086e-01 -8.76513273e-02 4.64914814e-02 3.04183438e-02
3.11213344e-01 -7.41904080e-01 3.95744950e-01 -4.01556939e-01
1.35716152e+00 -2.13387996e-01 -3.11106384e-01 -3.00786775e-02
-1.44249246e-01 2.54573971e-01 7.27464437e-01 1.02861536e+00
-1.42028213e-01 7.60110855e-01 2.78440416e-01 -9.28579092e-01
-1.27566469e+00 -9.23751593e-01 -2.70152152e-01 7.74964750e-01
3.79465893e-02 1.54707149e-01 -2.01469779e-01 -9.70139384e-01
3.57805610e-01 5.67307949e-01 -4.81356025e-01 -5.99840462e-01
-4.10283357e-01 -5.11111259e-01 -1.62188992e-01 7.34618723e-01
2.16908023e-01 -1.35393047e+00 -8.87853861e-01 3.68156224e-01
5.37199438e-01 -8.45778823e-01 9.31138620e-02 6.94160044e-01
-1.00617826e+00 -1.44584668e+00 -4.78991657e-01 -1.29674041e+00
1.09209096e+00 1.39050320e-01 1.33589768e+00 5.52085757e-01
-9.85760927e-01 5.01150191e-01 -4.55401450e-01 -9.46215987e-01
-9.25289333e-01 -3.94418776e-01 6.97480962e-02 -3.56419295e-01
4.49418068e-01 -9.35850106e-03 -4.91506428e-01 4.00093138e-01
-7.59447813e-01 -4.06901240e-01 7.97616124e-01 1.47295892e+00
3.31207842e-01 5.28646052e-01 8.88783395e-01 -9.11990285e-01
6.35434031e-01 -3.49019259e-01 -4.86781597e-01 1.28961727e-01
-6.04318202e-01 -2.62635887e-01 5.82244456e-01 -4.01973665e-01
-9.15698946e-01 4.61028609e-03 -7.67468214e-02 -5.59273481e-01
-5.48931479e-01 5.15310049e-01 2.24617377e-01 -1.57504424e-01
8.54552269e-01 -3.02394539e-01 2.51297563e-01 -5.80668747e-01
-2.31946215e-01 6.10653102e-01 2.09820673e-01 -4.37440425e-01
4.88224804e-01 3.74771863e-01 -6.21546023e-02 -9.34224725e-01
-8.81039441e-01 -4.19887930e-01 -9.01484311e-01 -4.33519840e-01
6.77783132e-01 -7.16145694e-01 -1.48763701e-01 5.71397305e-01
-1.11487174e+00 -3.76216710e-01 -5.45557559e-01 2.90975273e-01
-3.72510135e-01 2.87576288e-01 -7.95683563e-01 -8.09869111e-01
-6.19357266e-02 -1.24315417e+00 1.14872968e+00 2.55660325e-01
-3.11388392e-02 -9.20679510e-01 -4.13828373e-01 3.95634383e-01
1.54157609e-01 4.07484055e-01 1.13629985e+00 -5.82787216e-01
-4.18887705e-01 -6.87445164e-01 -8.45764428e-02 1.17576301e+00
6.47238076e-01 2.90846109e-01 -8.91631424e-01 -7.86249161e-01
1.51717365e-01 -6.79371536e-01 8.74996483e-01 3.24815154e-01
1.28236485e+00 1.03199929e-01 -4.37323123e-01 2.03584597e-01
1.34480762e+00 6.93117738e-01 4.76618439e-01 3.43594372e-01
6.93098664e-01 9.51715350e-01 1.28798485e+00 2.37916768e-01
-3.24365765e-01 3.07241023e-01 8.29243302e-01 -5.88597655e-01
5.09183062e-03 6.40280545e-02 1.42902225e-01 6.67228878e-01
2.31807441e-01 -1.05967196e-02 -1.09536707e+00 1.02625704e+00
-1.42943382e+00 -4.85285342e-01 -1.17016407e-02 1.97480381e+00
4.54117060e-01 6.06778026e-01 -1.95026189e-01 4.28566784e-01
6.19536340e-01 -4.57870394e-01 -7.24022090e-01 -5.66652060e-01
8.88955668e-02 3.74277949e-01 3.79458815e-01 -8.92325491e-02
-1.07323337e+00 5.59741259e-01 6.36495495e+00 1.93489894e-01
-9.28559542e-01 -3.99629511e-02 6.02702737e-01 1.11008465e-01
2.43997425e-01 -2.17306510e-01 -5.83409369e-01 2.86723226e-01
5.79145253e-01 7.77817488e-01 -5.26369885e-02 1.05283844e+00
1.31299675e-01 -2.85528243e-01 -1.66241479e+00 7.52277851e-01
2.86515623e-01 -1.24843740e+00 -2.57703185e-01 -1.04548365e-01
9.37002957e-01 -2.05151126e-01 1.59553006e-01 4.43470091e-01
3.79897326e-01 -1.02076423e+00 4.21762019e-01 3.00320119e-01
4.76160258e-01 -6.68174088e-01 1.15695524e+00 1.35744244e-01
-6.00155413e-01 -7.76366949e-01 -7.93472886e-01 -2.14583874e-01
-7.41172640e-04 6.95079386e-01 -1.43975914e+00 3.06286126e-01
1.11179364e+00 8.20949018e-01 -7.35197484e-01 1.25325572e+00
-1.28684789e-01 4.98819590e-01 4.57332768e-02 1.24860108e-01
1.15548238e-01 2.39472136e-01 1.94527969e-01 8.89363766e-01
4.32103336e-01 -7.06078649e-01 4.93559778e-01 9.12040472e-01
3.44254106e-01 -5.15669107e-01 -1.05063927e+00 1.99225522e-03
2.13384405e-01 9.34309244e-01 -7.61941373e-01 7.62504712e-02
-4.93922263e-01 8.61961663e-01 1.82984948e-01 1.84468433e-01
-3.12222868e-01 -4.18276966e-01 6.35013402e-01 2.20843777e-01
5.73285818e-01 -1.83115453e-01 -2.27784917e-01 -4.95216578e-01
-2.05151681e-02 -1.14949405e+00 2.43522167e-01 -8.48717391e-01
-1.68338764e+00 2.16717422e-01 -1.92341343e-01 -1.25066555e+00
3.80794145e-02 -1.30605829e+00 -6.71225786e-01 4.02503878e-01
-1.25882423e+00 -1.02556336e+00 -4.25114244e-01 9.32359546e-02
1.13720024e+00 -2.43283629e-01 5.46248198e-01 2.29846686e-01
-4.40169901e-01 3.81111592e-01 1.28442079e-01 1.80794969e-01
6.53468192e-01 -1.28786230e+00 3.88517052e-01 8.30117524e-01
1.31752670e-01 4.01530743e-01 7.96940565e-01 -9.24405813e-01
-1.32338202e+00 -1.14761341e+00 6.50386885e-02 -5.65019310e-01
2.65329808e-01 -3.81316483e-01 -1.22963059e+00 6.22367501e-01
-1.19332768e-01 1.47191986e-01 2.63615012e-01 3.01115036e-01
7.04311766e-03 -2.49807298e-01 -1.47356772e+00 2.09880874e-01
8.34931493e-01 -4.53329951e-01 -5.62921762e-01 7.29949117e-01
6.67917907e-01 -3.83856505e-01 -1.11157703e+00 6.71643198e-01
1.81007683e-01 -9.70361114e-01 8.39620054e-01 -5.86258411e-01
6.01403475e-01 -1.90483317e-01 1.17417358e-01 -1.72626829e+00
-3.58119160e-01 -4.52392325e-02 2.25444306e-02 1.09027314e+00
4.78209794e-01 -4.49493468e-01 6.75646007e-01 2.41943583e-01
-8.49065185e-01 -8.32505465e-01 -5.85960209e-01 -1.08186829e+00
9.83343110e-04 -2.55507499e-01 4.93304819e-01 7.60219216e-01
-3.17025721e-01 -1.05658977e-03 -2.05999568e-01 4.49363649e-01
2.89820552e-01 -3.63120623e-02 5.70498407e-01 -1.44519985e+00
-2.58810908e-01 1.01932876e-01 -8.91010702e-01 -6.61234379e-01
-2.33875558e-01 -4.75583792e-01 6.28890276e-01 -1.75001955e+00
-2.75474072e-01 -6.68583512e-01 -2.68221885e-01 5.32002807e-01
-7.92168006e-02 -1.12019042e-02 1.10916577e-01 -1.25484020e-01
-3.16869974e-01 2.20320120e-01 1.59424424e+00 -7.08978832e-01
-3.85041058e-05 2.33990833e-01 -4.18656468e-01 6.23557210e-01
8.05105150e-01 -3.60532522e-01 -3.99823666e-01 -7.08592713e-01
-1.25369996e-01 -2.35388860e-01 3.58580083e-01 -1.36661267e+00
-9.33449641e-02 2.53495961e-01 8.94968092e-01 -6.42761111e-01
3.65951002e-01 -1.01228583e+00 -3.28617394e-01 7.35113740e-01
-8.47743675e-02 5.27744032e-02 5.19588113e-01 7.01787353e-01
-1.92474872e-01 -7.98670828e-01 7.15664506e-01 -6.45177543e-01
-1.08326912e+00 2.22096182e-02 -3.98429513e-01 -2.44838402e-01
1.06071866e+00 -6.37774765e-01 -2.84395874e-01 6.71152622e-02
-7.01475561e-01 1.67941481e-01 6.73972368e-01 7.83020973e-01
1.03032970e+00 -9.62209225e-01 -5.90370119e-01 5.46056867e-01
4.08186018e-01 3.60978007e-01 3.50223511e-01 3.37573707e-01
-6.94064558e-01 1.13100164e-01 -4.76786375e-01 -1.14841831e+00
-1.12490118e+00 8.53552163e-01 3.43422800e-01 -1.65413655e-02
-6.56415641e-01 1.12880015e+00 2.00193986e-01 -5.59397340e-01
2.54114896e-01 -7.39692628e-01 1.02758706e-02 -1.46401182e-01
3.34585488e-01 2.02358171e-01 5.66281378e-01 1.64731313e-02
4.70549390e-02 3.06105167e-01 -5.70271730e-01 7.75319338e-01
1.40707815e+00 2.19104245e-01 2.35505238e-01 7.00890124e-01
9.58438218e-01 -5.93930066e-01 -1.81297576e+00 -4.63118628e-02
1.37091100e-01 -6.75947249e-01 1.11165613e-01 -1.01815057e+00
-1.31989920e+00 1.02124858e+00 1.18212867e+00 5.76149285e-01
9.56086099e-01 8.81234333e-02 6.10450208e-01 4.72950190e-01
5.37167609e-01 -1.38847613e+00 7.42104292e-01 2.63995141e-01
1.02227545e+00 -1.82295418e+00 -9.05972049e-02 -3.51508290e-01
-4.20099229e-01 1.41864443e+00 1.14699090e+00 -1.79015353e-01
3.80790532e-01 2.79155254e-01 1.50366858e-01 -8.93617034e-01
-6.14876568e-01 -6.63258061e-02 5.76562472e-02 1.38211346e+00
1.12583794e-01 -1.58073425e-01 5.43172359e-01 8.19635093e-02
2.14146629e-01 -2.37225741e-01 6.93454206e-01 1.46867228e+00
-6.54695272e-01 -8.62304032e-01 -5.75452387e-01 1.13546753e+00
-2.96303660e-01 3.14754009e-01 -1.08103536e-01 8.89940798e-01
6.23976231e-01 1.29874063e+00 3.08288902e-01 -4.96669978e-01
6.67744994e-01 -2.66999424e-01 7.22274959e-01 -1.37583089e+00
-5.09234786e-01 -3.80307376e-01 1.96338698e-01 -2.76419491e-01
2.65062675e-02 -8.90484929e-01 -1.01125634e+00 3.39205980e-01
-7.75841832e-01 -3.90522867e-01 7.11439788e-01 8.98205996e-01
1.45230144e-01 1.16262698e+00 6.35096610e-01 -1.10757041e+00
-7.30295420e-01 -1.08307362e+00 -7.47069418e-01 6.14050448e-01
6.41483009e-01 -9.42761004e-01 -3.57988805e-01 1.85277969e-01] | [7.384544372558594, 1.9597846269607544] |
f718e62a-08e2-44fc-a74e-c65e398814c7 | gental-generative-denoising-skip-gram | null | null | https://openreview.net/forum?id=36SHWj0Gp1 | https://openreview.net/pdf?id=36SHWj0Gp1 | GenTAL: Generative Denoising Skip-gram Transformer for Unsupervised Binary Code Similarity Detection | Binary code similarity detection serves a critical role in cybersecurity. It alleviates the huge manual effort required in the reverse engineering process for malware analysis and vulnerability detection, where often the original source code is not available for analysis. Most of the existing solutions focus on a manual feature engineering process and customized code matching algorithms that are inefficient and inaccurate. Recent deep-learning-based solutions embed the semantics of binary code into a latent space through supervised contrastive learning. However, one cannot cover all its possible forms in the training set to learn the variance of the same semantic. In this paper, we propose an unsupervised model aiming to learn the intrinsic representation of assembly code semantics. Specifically, we propose a Transformer-based auto-encoder-like language model for the low-level assembly code grammar to capture the abstract semantic representation. By coupling a Transformer encoder and a skip-gram-style loss design, it can learn a compact representation that is robust against different compilation options. We conduct experiments on four different block-level code similarity tasks. It shows that our method is more robust compared to the state-of-the-art. | ['Christopher James Molloy', 'Hanbo Yu', 'Philippe Charland', 'Steven Ding', 'Litao Li'] | 2021-09-29 | null | null | null | null | ['vulnerability-detection'] | ['miscellaneous'] | [ 2.25701138e-01 -4.12780493e-01 -4.30699646e-01 -5.20236731e-01
-7.89480269e-01 -7.44151950e-01 4.08839077e-01 4.67777431e-01
-9.64747593e-02 4.17716466e-02 1.99336000e-02 -8.18381846e-01
9.58150774e-02 -8.10052276e-01 -7.65452147e-01 -4.48369533e-01
1.35442009e-03 9.22259241e-02 2.77666688e-01 -2.14837000e-01
5.19541264e-01 1.43014580e-01 -1.36427820e+00 5.05911171e-01
1.04208195e+00 8.70616257e-01 3.84591013e-01 4.49165076e-01
-3.07687402e-01 8.75974119e-01 -5.43146729e-01 -7.30292261e-01
2.83646494e-01 -3.01747173e-01 -5.60892582e-01 -1.87177181e-01
9.05304328e-02 -2.00722739e-01 -3.78033966e-01 1.53787863e+00
-4.91019115e-02 -4.29655135e-01 6.61556542e-01 -1.31600201e+00
-6.10841334e-01 6.36294901e-01 -5.48997641e-01 -1.01518696e-02
2.64802247e-01 1.60487011e-01 1.11576414e+00 -4.27398980e-01
3.35971028e-01 1.17557085e+00 8.65022063e-01 2.91639477e-01
-1.10075545e+00 -6.43595815e-01 5.12830392e-02 2.69148678e-01
-1.23033381e+00 -8.87344182e-02 1.03615212e+00 -6.52965844e-01
1.03528523e+00 1.00655235e-01 3.31905931e-01 1.35408461e+00
6.45806372e-01 5.59693098e-01 7.79231548e-01 -2.04563320e-01
2.29284748e-01 2.39368528e-01 4.39123780e-01 8.91291499e-01
3.58339280e-01 1.92145273e-01 7.63682276e-02 -5.77044487e-01
2.24582911e-01 5.33779979e-01 -1.18617266e-01 -6.99157894e-01
-6.39215052e-01 1.19398677e+00 4.81092483e-01 4.41746056e-01
1.01071671e-01 1.85731083e-01 8.97088408e-01 5.72111845e-01
2.94244736e-01 4.24581647e-01 -5.43109477e-01 -2.23920569e-01
-9.79052842e-01 -3.83184068e-02 7.16294825e-01 9.05618250e-01
8.21934342e-01 5.50239682e-02 1.11631572e-01 5.38352847e-01
5.09826958e-01 2.25500122e-01 6.81996584e-01 -2.20752820e-01
5.67920029e-01 9.42427516e-01 -4.99574900e-01 -1.33154488e+00
-2.00446695e-02 -6.75892055e-01 -5.87431908e-01 1.60450771e-01
4.46713343e-02 2.60752469e-01 -5.29979527e-01 1.72133470e+00
-7.44913742e-02 1.18331186e-01 1.37098609e-02 3.26551795e-01
1.71429485e-01 5.60909927e-01 -1.39740407e-01 1.15093865e-01
1.33029521e+00 -9.15269434e-01 -3.75258297e-01 -4.95978475e-01
9.07696009e-01 -7.04667330e-01 1.26572335e+00 2.84705579e-01
-4.10487950e-01 -4.33811009e-01 -1.37029755e+00 1.72777191e-01
-4.00460124e-01 1.15645245e-01 7.01907396e-01 1.09260380e+00
-6.17299259e-01 6.87161744e-01 -7.39656150e-01 -9.13079679e-02
2.90753961e-01 1.18314229e-01 -3.66070509e-01 8.48835520e-03
-1.06207609e+00 6.38020992e-01 6.21957302e-01 -2.33749479e-01
-1.12329853e+00 -5.75397313e-01 -1.17567456e+00 2.39102826e-01
3.51994127e-01 -3.26506525e-01 9.30476069e-01 -1.06434464e+00
-1.38431358e+00 7.32773244e-01 7.26268217e-02 -4.95807439e-01
1.55001208e-01 -1.16059281e-01 -3.48822594e-01 -1.35190085e-01
-7.77406096e-02 -1.47723779e-01 1.43719566e+00 -1.08996081e+00
-2.69052923e-01 -2.90447503e-01 2.18535945e-01 -4.39060748e-01
-6.34310603e-01 2.96005458e-01 -9.78448391e-02 -9.89098489e-01
-2.76713669e-01 -8.58119309e-01 -1.44983381e-02 -1.72529265e-01
-2.62694985e-01 1.20128751e-01 1.06704426e+00 -9.75686610e-01
1.48573422e+00 -2.45610547e+00 2.22198665e-01 1.55858472e-01
1.88884526e-01 4.10997450e-01 -9.21526775e-02 4.63998646e-01
-2.28945881e-01 1.03684723e-01 -7.07611978e-01 -2.89020836e-01
1.88024268e-01 -8.18519965e-02 -8.12195480e-01 6.21994972e-01
1.83533549e-01 8.24606359e-01 -8.44825923e-01 -2.89002955e-01
-1.58151254e-01 3.12022209e-01 -8.96004736e-01 3.97943467e-01
-3.60452056e-01 2.85840295e-02 -3.46309096e-01 6.40208781e-01
6.04672909e-01 -2.25220487e-01 2.20872521e-01 -1.19571067e-01
1.22423254e-01 5.84990382e-01 -7.19953418e-01 1.94114053e+00
-7.97555566e-01 4.82906103e-01 -1.53165579e-01 -1.20346570e+00
1.13764584e+00 -7.80164823e-02 -1.09134600e-01 -3.20037782e-01
3.06654155e-01 4.76657778e-01 1.28672451e-01 -3.33649755e-01
2.37627998e-01 5.35514541e-02 -5.24143815e-01 6.08470857e-01
-3.96097898e-02 -1.59660220e-01 -2.39573523e-01 1.30780146e-01
1.28638124e+00 3.84221263e-02 4.33021069e-01 -2.56682754e-01
8.43332350e-01 -2.26261169e-01 5.39399743e-01 5.16851008e-01
1.86044052e-02 7.74892271e-02 7.58539379e-01 -3.72211665e-01
-1.12500823e+00 -9.40194845e-01 3.40618975e-02 7.67808139e-01
4.64882292e-02 -8.73322070e-01 -9.62601125e-01 -1.07448554e+00
-1.27042338e-01 7.05252469e-01 -4.46222991e-01 -8.15090775e-01
-5.89987278e-01 -5.35516679e-01 7.69236386e-01 3.79951060e-01
3.74401540e-01 -6.30753040e-01 -5.93873441e-01 7.77646387e-03
1.11391433e-02 -8.12629700e-01 -7.20284820e-01 2.99935162e-01
-7.75428057e-01 -1.19150639e+00 -8.77337381e-02 -8.20924222e-01
6.32847667e-01 1.93485916e-02 8.44336867e-01 2.57949382e-01
-4.73956347e-01 -1.79561570e-01 -5.55545568e-01 9.08978060e-02
-8.59406888e-01 1.93873748e-01 -1.05052002e-01 4.34698611e-02
4.09517676e-01 -5.90307593e-01 -1.53904855e-01 -2.70807091e-02
-9.17511523e-01 -2.55232900e-01 5.36499679e-01 1.02686036e+00
3.18510272e-02 3.46589565e-01 1.98877871e-01 -7.84706712e-01
6.32838726e-01 -6.13973975e-01 -8.34574044e-01 3.59906346e-01
-5.19242764e-01 5.29240370e-01 1.17274952e+00 -3.54621470e-01
-8.19625974e-01 -4.54105111e-03 -2.05333054e-01 -5.26530147e-01
1.01037554e-01 3.50915968e-01 -4.92026478e-01 -1.60314471e-01
4.97098684e-01 5.43939531e-01 8.50397348e-02 -5.24838984e-01
1.91471025e-01 8.77793431e-01 2.56480515e-01 -5.31941354e-01
1.19431889e+00 1.40696451e-01 -1.74357057e-01 -3.84807289e-01
-5.19093633e-01 -2.13509038e-01 -4.20524240e-01 4.49229419e-01
6.44097567e-01 -5.92734933e-01 -3.84330392e-01 5.81100404e-01
-1.36476517e+00 -1.27857983e-01 7.00028166e-02 1.00586191e-01
-5.04144967e-01 9.26160574e-01 -6.33334816e-01 -4.60645646e-01
-3.10669094e-01 -1.76542997e+00 1.10236275e+00 -2.84295887e-01
-8.74360874e-02 -9.30231929e-01 2.58357763e-01 -4.32997905e-02
4.71007138e-01 3.34622040e-02 1.54450810e+00 -8.21320236e-01
-4.99200583e-01 -4.25507009e-01 -1.56006142e-01 4.61845398e-01
3.30425501e-01 -1.93684679e-02 -8.49046707e-01 -5.10147929e-01
5.20005226e-01 -2.41347745e-01 8.04742992e-01 -2.10774243e-01
1.47281635e+00 -5.43130398e-01 -3.50478709e-01 9.38961089e-01
1.55582988e+00 2.98234463e-01 5.68023682e-01 2.98930347e-01
7.42866039e-01 5.28138816e-01 4.08615917e-01 3.07418138e-01
1.14274941e-01 8.71860445e-01 6.64948523e-01 3.93755138e-01
4.07254130e-01 -4.47495461e-01 7.02412546e-01 9.31600988e-01
7.36900270e-01 8.14842246e-03 -9.69731092e-01 3.04276675e-01
-1.61379623e+00 -9.56748903e-01 2.10008293e-01 2.21315074e+00
8.82050931e-01 2.48104855e-01 -7.55468383e-02 3.98778617e-01
7.40390360e-01 2.46827856e-01 -3.25384796e-01 -6.25025094e-01
2.82935739e-01 2.19086409e-01 3.85731786e-01 4.80000556e-01
-1.24563074e+00 8.01188231e-01 5.58012247e+00 1.31488764e+00
-1.12946498e+00 3.32873821e-01 3.61925721e-01 5.06895483e-01
-5.95838130e-01 4.36604857e-01 -6.07881725e-01 8.54867995e-01
1.09614789e+00 -9.36375260e-02 3.91311973e-01 1.21788144e+00
-3.90974075e-01 4.72256690e-01 -1.16689038e+00 1.04374290e+00
2.31296554e-01 -1.16083133e+00 1.60713326e-02 -2.64487248e-02
3.22207391e-01 -2.71562457e-01 1.51683301e-01 4.23694164e-01
-1.37180672e-03 -9.11035478e-01 8.30866039e-01 2.41174296e-01
7.21721053e-01 -8.54985178e-01 7.77953207e-01 3.92761379e-01
-1.38343871e+00 -4.63023424e-01 -3.82645071e-01 1.76532567e-01
-1.87204346e-01 5.35676420e-01 -6.19341016e-01 5.97579479e-01
3.40558469e-01 7.97371268e-01 -7.53523827e-01 6.96022570e-01
-2.83049315e-01 4.34557498e-01 2.30891511e-01 -3.28336805e-02
3.00594389e-01 -1.59427468e-02 5.24940550e-01 1.32035649e+00
3.20905447e-01 -6.17770195e-01 3.11193556e-01 1.09975159e+00
5.40388562e-02 -9.13520753e-02 -8.42696428e-01 -4.90007102e-01
4.35106248e-01 9.66534972e-01 -6.25230491e-01 -1.61070935e-02
-4.63998199e-01 1.09961879e+00 3.21105540e-01 -1.28407255e-01
-1.08326435e+00 -8.46596897e-01 8.29470992e-01 -1.00112222e-01
4.28105772e-01 -4.33390230e-01 -2.36105084e-01 -1.36649907e+00
9.40617099e-02 -1.31424654e+00 2.21572474e-01 -2.45023951e-01
-1.24929750e+00 9.06718850e-01 -4.56763506e-02 -1.36478126e+00
-5.31697154e-01 -8.01926017e-01 -7.73813903e-01 6.48449361e-01
-1.29665101e+00 -1.00087142e+00 -4.11568061e-02 4.03585464e-01
5.65326035e-01 -6.64148629e-01 8.58650267e-01 4.38981324e-01
-5.41554391e-01 9.96130347e-01 1.85606167e-01 2.29758367e-01
4.80808794e-01 -9.71595645e-01 8.11091304e-01 1.15047669e+00
3.54772061e-02 1.08123815e+00 6.06611431e-01 -7.81040370e-01
-1.60333180e+00 -1.08727074e+00 7.35736132e-01 -3.56654316e-01
1.04985070e+00 -7.52206266e-01 -9.79843795e-01 5.46905994e-01
-1.06224962e-01 -1.99300170e-01 7.38344610e-01 -3.00543964e-01
-1.20745313e+00 -4.65700738e-02 -1.11270630e+00 4.25781041e-01
7.83034325e-01 -1.09752357e+00 -6.55812681e-01 9.91864279e-02
1.18350327e+00 1.75320320e-02 -4.64007080e-01 2.10572273e-01
4.40168679e-01 -8.93100560e-01 8.51448238e-01 -6.20886445e-01
5.95201433e-01 -3.32365394e-01 -3.25187355e-01 -1.07843792e+00
-3.30529749e-01 -5.44660091e-01 -2.61898249e-01 1.12352109e+00
1.10072307e-01 -6.00769699e-01 5.39227128e-01 -9.46518481e-02
-1.48076177e-01 -6.60124540e-01 -7.92284191e-01 -1.03278303e+00
-4.87948060e-02 -4.26810890e-01 7.93272793e-01 9.79033470e-01
1.23457901e-01 1.36485323e-02 -2.84696370e-01 1.13834053e-01
7.65528560e-01 4.25165504e-01 5.76371312e-01 -1.04891038e+00
-8.71846020e-01 -5.95068157e-01 -8.07421386e-01 -7.90886760e-01
6.36501253e-01 -1.24138963e+00 -9.44111869e-02 -8.16020608e-01
4.09658790e-01 -3.10315788e-01 -1.87863842e-01 4.79627669e-01
-4.43072878e-02 -1.70409322e-01 1.28358864e-04 1.04994267e-01
-1.67699054e-01 6.49013698e-01 3.71886432e-01 -5.66151738e-01
2.14802444e-01 -1.81433067e-01 -6.73203468e-01 6.59525514e-01
9.21178997e-01 -8.37886691e-01 -6.54913723e-01 -3.58847082e-01
2.92287380e-01 -1.45200402e-01 4.84051555e-01 -8.25278342e-01
5.23841530e-02 -9.21916440e-02 -3.24256897e-01 -2.64231473e-01
3.41887698e-02 -1.00501776e+00 1.52297258e-01 9.51942921e-01
-2.27178559e-01 2.43212789e-01 1.54745474e-01 5.90597749e-01
-2.51867086e-01 -7.52395213e-01 7.82283545e-01 -7.86957666e-02
-4.88004625e-01 3.31665576e-01 -3.26410353e-01 1.04983803e-02
9.33643222e-01 -1.10392228e-01 -3.01796436e-01 -6.12101778e-02
-2.11122874e-02 -3.23277295e-01 9.11379814e-01 6.91405058e-01
7.22094178e-01 -1.26640379e+00 -4.40276653e-01 6.03321314e-01
3.79073054e-01 -7.10640430e-01 -4.12082598e-02 5.09059310e-01
-5.33371449e-01 4.12462562e-01 -4.13723290e-01 -3.43616009e-01
-1.22614312e+00 9.99228537e-01 1.75652906e-01 -4.72021013e-01
-5.37709475e-01 6.25783801e-01 1.78507879e-01 -5.38311601e-01
2.16537580e-01 -2.36521304e-01 3.75545546e-02 -3.11504781e-01
5.85977376e-01 9.10931155e-02 5.71157970e-02 -5.62813103e-01
-5.38760066e-01 6.12767458e-01 -2.57697374e-01 3.54408801e-01
1.24759424e+00 1.93577588e-01 -5.15720844e-01 1.90621734e-01
1.75245810e+00 3.41309682e-02 -1.00800955e+00 -2.67610345e-02
3.96877527e-01 -7.38110065e-01 -7.23449513e-02 -2.83831716e-01
-9.02077734e-01 1.16009223e+00 5.44134617e-01 4.93741333e-02
1.06102407e+00 -1.75909162e-01 8.78804147e-01 5.16023397e-01
6.76147997e-01 -6.17911160e-01 2.66304076e-01 7.33209252e-01
5.64664900e-01 -1.04823792e+00 -3.36429298e-01 -4.33599144e-01
-2.48939216e-01 1.25767612e+00 4.28435266e-01 -2.57040352e-01
7.04488575e-01 5.76357007e-01 -4.71353978e-01 -1.10536814e-01
-5.87504685e-01 2.43580207e-01 2.61818945e-01 4.83369082e-01
3.41896325e-01 -5.76711372e-02 -2.44881853e-01 8.26609790e-01
-1.35099500e-01 -5.63919246e-01 3.87161732e-01 1.02732289e+00
-3.98550838e-01 -1.58574331e+00 -2.11866766e-01 2.97244430e-01
-4.21015859e-01 -3.67541760e-01 -3.97886872e-01 3.31395507e-01
6.86844736e-02 7.93001235e-01 -1.36686146e-01 -8.50881457e-01
-4.67905030e-02 1.45347014e-01 4.53234762e-01 -8.04861128e-01
-6.66590154e-01 -3.36166054e-01 -3.13111037e-01 -5.93695283e-01
1.02549039e-01 -3.87431592e-01 -9.80314791e-01 -1.44958168e-01
-2.48732269e-01 1.33827761e-01 6.22943044e-01 7.52659559e-01
3.36490989e-01 5.45897305e-01 9.09245849e-01 -5.29999435e-01
-9.85280156e-01 -5.50851464e-01 -3.46807361e-01 4.14391279e-01
4.22561198e-01 -7.27329016e-01 -4.77213770e-01 -4.65700887e-02] | [7.127371788024902, 7.808948993682861] |
f05543c7-00bf-4888-85cf-19eea67f96ca | dit-3d-exploring-plain-diffusion-transformers | 2307.01831 | null | https://arxiv.org/abs/2307.01831v1 | https://arxiv.org/pdf/2307.01831v1.pdf | DiT-3D: Exploring Plain Diffusion Transformers for 3D Shape Generation | Recent Diffusion Transformers (e.g., DiT) have demonstrated their powerful effectiveness in generating high-quality 2D images. However, it is still being determined whether the Transformer architecture performs equally well in 3D shape generation, as previous 3D diffusion methods mostly adopted the U-Net architecture. To bridge this gap, we propose a novel Diffusion Transformer for 3D shape generation, namely DiT-3D, which can directly operate the denoising process on voxelized point clouds using plain Transformers. Compared to existing U-Net approaches, our DiT-3D is more scalable in model size and produces much higher quality generations. Specifically, the DiT-3D adopts the design philosophy of DiT but modifies it by incorporating 3D positional and patch embeddings to adaptively aggregate input from voxelized point clouds. To reduce the computational cost of self-attention in 3D shape generation, we incorporate 3D window attention into Transformer blocks, as the increased 3D token length resulting from the additional dimension of voxels can lead to high computation. Finally, linear and devoxelization layers are used to predict the denoised point clouds. In addition, our transformer architecture supports efficient fine-tuning from 2D to 3D, where the pre-trained DiT-2D checkpoint on ImageNet can significantly improve DiT-3D on ShapeNet. Experimental results on the ShapeNet dataset demonstrate that the proposed DiT-3D achieves state-of-the-art performance in high-fidelity and diverse 3D point cloud generation. In particular, our DiT-3D decreases the 1-Nearest Neighbor Accuracy of the state-of-the-art method by 4.59 and increases the Coverage metric by 3.51 when evaluated on Chamfer Distance. | ['Zhenguo Li', 'Matthias Nießner', 'Lanqing Hong', 'Lewei Yao', 'Ruihang Chu', 'Enze Xie', 'Shentong Mo'] | 2023-07-04 | null | null | null | null | ['3d-shape-generation', 'point-cloud-generation', 'philosophy'] | ['computer-vision', 'computer-vision', 'miscellaneous'] | [-1.38535142e-01 2.08437100e-01 3.24858129e-01 -1.39363781e-01
-9.98453796e-01 -3.73483360e-01 6.24192536e-01 -2.95137912e-01
1.11079134e-01 4.13457066e-01 2.26801053e-01 -2.88435936e-01
5.21262214e-02 -1.49504578e+00 -1.14740586e+00 -5.67650974e-01
1.79696187e-01 7.10845768e-01 2.20567912e-01 -4.08236295e-01
2.88307127e-02 9.75161314e-01 -1.48115838e+00 1.36211380e-01
1.08252978e+00 1.30029738e+00 1.39217958e-01 2.36639902e-01
-3.93096209e-01 1.15568683e-01 -6.15306139e-01 -3.91452938e-01
5.70876718e-01 -1.96087044e-02 -3.88288200e-01 4.89423499e-02
6.83017254e-01 -6.35636032e-01 -3.67239535e-01 6.39412105e-01
8.69674563e-01 -5.08342087e-02 8.91268849e-01 -1.12731552e+00
-1.24516010e+00 4.89445418e-01 -5.84385633e-01 -2.04484224e-01
-7.52766756e-03 5.52705288e-01 8.20613265e-01 -1.37902844e+00
5.68480611e-01 1.55597258e+00 8.82332325e-01 5.68906605e-01
-1.22510672e+00 -8.43954504e-01 1.21011607e-01 -2.53393024e-01
-1.49145842e+00 7.22019523e-02 1.01816952e+00 -3.98593694e-01
1.35039055e+00 1.42351523e-01 1.04031491e+00 1.04889905e+00
9.92212072e-02 7.41315663e-01 6.83117449e-01 5.21530397e-02
2.54123062e-01 -4.32100415e-01 -5.20785391e-01 5.57175934e-01
1.93518952e-01 3.17682892e-01 -2.07667455e-01 -1.13533765e-01
1.44736779e+00 -2.45082211e-02 -7.70845860e-02 -4.01287168e-01
-1.09713888e+00 7.47899652e-01 1.04801726e+00 1.44044593e-01
-6.13078117e-01 4.15931821e-01 1.11710191e-01 7.34968632e-02
9.87292886e-01 4.45066869e-01 -3.99097204e-01 -1.11518595e-02
-9.46820855e-01 6.10864222e-01 2.93938637e-01 1.14756870e+00
6.64608717e-01 4.63984549e-01 -3.47230315e-01 7.55953908e-01
2.60634005e-01 8.50395441e-01 1.62462652e-01 -8.27451825e-01
5.14931202e-01 8.94044638e-01 -1.34935006e-01 -7.37761676e-01
-1.43796578e-01 -5.05099416e-01 -1.18596721e+00 4.99702960e-01
-1.35628641e-01 4.21118177e-02 -1.51407397e+00 1.38972127e+00
5.53563476e-01 3.14798027e-01 -1.53801993e-01 9.16414380e-01
1.11601305e+00 7.22998619e-01 -2.30473846e-01 3.95308971e-01
9.01676774e-01 -7.85338461e-01 -1.75808430e-01 2.60880366e-02
1.72664821e-01 -7.95084298e-01 1.16621470e+00 5.98456189e-02
-1.33504939e+00 -6.84027433e-01 -8.65955651e-01 -2.86799133e-01
-2.24025518e-01 -2.92209595e-01 5.62227547e-01 4.29042637e-01
-1.20124769e+00 6.26737535e-01 -9.69307125e-01 4.32740971e-02
9.22462106e-01 2.82267243e-01 -1.12724029e-01 -2.55386770e-01
-1.01109922e+00 5.65989792e-01 -6.31193221e-02 6.45162072e-04
-9.56055641e-01 -1.26581252e+00 -9.06837285e-01 1.89505473e-01
-1.12646252e-01 -1.29237068e+00 1.18368506e+00 -1.60805017e-01
-1.43804276e+00 5.82648337e-01 -9.30647403e-02 -3.48256826e-01
6.82148159e-01 3.51337306e-02 1.27707645e-01 -7.48670921e-02
2.34815344e-01 1.25637972e+00 8.62366617e-01 -1.43256903e+00
-3.19567472e-01 -4.06028301e-01 6.41801581e-02 3.19833279e-01
-1.21062130e-01 -7.73301601e-01 -4.93139088e-01 -1.05406117e+00
3.69849175e-01 -6.51503325e-01 -4.42435831e-01 2.67604113e-01
-3.29552948e-01 -4.54866976e-01 9.50518012e-01 -2.97827631e-01
7.50243604e-01 -1.98270369e+00 1.05298020e-01 2.23477542e-01
3.97804677e-01 1.70550913e-01 -3.38889539e-01 1.87301278e-01
-2.01532547e-03 4.97389257e-01 -4.77914333e-01 -5.46418548e-01
1.62297159e-01 2.54778236e-01 -3.61267895e-01 2.72264015e-02
5.71449041e-01 1.33775628e+00 -7.61717975e-01 -2.16962397e-01
4.78288382e-01 9.35869038e-01 -8.94711912e-01 9.99025330e-02
-4.95581031e-01 3.23164016e-01 -6.07339442e-01 8.19352210e-01
1.03635049e+00 -3.22299927e-01 -6.06751800e-01 -3.67564231e-01
-4.71725091e-02 2.25412026e-01 -6.95415378e-01 1.91250515e+00
-5.96758664e-01 1.37484983e-01 -2.26590469e-01 -4.42290634e-01
1.14519107e+00 2.67697483e-01 6.25791371e-01 -7.28275776e-01
3.83799598e-02 2.63708562e-01 -2.43125975e-01 3.00320648e-02
5.67441046e-01 -1.62433818e-01 8.05249959e-02 2.79146016e-01
-1.49986044e-01 -9.79567647e-01 -1.62187740e-01 8.29070285e-02
9.34877694e-01 2.38603577e-01 -3.88911605e-01 1.70830619e-02
6.48657158e-02 -4.13805097e-02 4.44887578e-01 4.78751421e-01
3.49442482e-01 1.10922694e+00 2.62083977e-01 -4.98425663e-01
-1.37745965e+00 -1.38235307e+00 -1.52583882e-01 1.54138461e-01
1.16510980e-01 -2.13337138e-01 -7.64632106e-01 -6.57433569e-01
4.97723132e-01 8.07095408e-01 -5.74484944e-01 -7.38386512e-02
-5.69401443e-01 -5.51268518e-01 4.98751819e-01 7.20699370e-01
5.98778784e-01 -9.88960207e-01 -2.00559065e-01 2.84776241e-01
3.41305323e-02 -8.24024022e-01 -5.84172428e-01 -1.27308339e-01
-1.27963865e+00 -8.03820848e-01 -9.59379494e-01 -6.44402921e-01
9.07113969e-01 2.76568681e-01 1.27188694e+00 3.05527207e-02
3.48076690e-04 1.86106786e-01 -3.49338025e-01 -4.44602519e-01
-2.69344330e-01 2.68944979e-01 -1.33942425e-01 -3.65344316e-01
1.84768900e-01 -9.61655736e-01 -8.18989277e-01 2.47622311e-01
-1.16602445e+00 7.40039870e-02 6.21624529e-01 8.59908640e-01
1.09363365e+00 -2.14957967e-02 5.37552953e-01 -4.91594255e-01
6.83324814e-01 -3.38378310e-01 -4.97021854e-01 -1.65077761e-01
-5.93932569e-01 7.42349848e-02 5.54146647e-01 -2.57856578e-01
-1.00767720e+00 -1.11187011e-01 -7.77296782e-01 -1.03414392e+00
6.39858171e-02 1.85840964e-01 -1.91497073e-01 -2.53248483e-01
6.57556534e-01 3.07635844e-01 2.31148005e-02 -6.61760032e-01
5.84078312e-01 3.83800507e-01 2.41880521e-01 -6.46761894e-01
1.14699626e+00 6.42308593e-01 -2.49043200e-02 -7.20512569e-01
-5.73829234e-01 1.00479156e-01 -4.40063030e-01 -1.22387931e-01
8.43184531e-01 -1.01314282e+00 -3.84499490e-01 7.01891720e-01
-1.46765804e+00 -4.75511789e-01 -7.21436739e-01 -1.88627765e-02
-5.63343346e-01 1.59215674e-01 -5.86964548e-01 -3.81248087e-01
-7.75790453e-01 -1.43308175e+00 1.58337891e+00 -1.00131325e-01
1.36484662e-02 -6.63028777e-01 -2.77506709e-01 3.13753098e-01
5.80929279e-01 4.64471608e-01 1.03466427e+00 1.79414041e-02
-1.10001230e+00 -3.89925428e-02 -3.58143330e-01 5.73672533e-01
2.30891287e-01 -1.97213009e-01 -8.58731568e-01 -1.35099620e-01
-1.31822497e-01 -1.32055685e-01 7.63704956e-01 5.72557867e-01
1.36966360e+00 -2.07101211e-01 -2.30293602e-01 9.73000228e-01
1.29580152e+00 4.38512601e-02 7.62811422e-01 3.53705473e-02
9.97135580e-01 -6.48214221e-02 2.70180076e-01 3.53176117e-01
5.89897752e-01 5.04696429e-01 8.13123167e-01 -3.47854912e-01
-6.83362901e-01 -6.63865447e-01 3.65047194e-02 1.04898524e+00
-6.66871965e-02 -2.05795899e-01 -8.31903756e-01 7.07174838e-01
-1.49537897e+00 -5.92253149e-01 -8.33322704e-02 1.97831690e+00
7.72988260e-01 2.22090021e-01 -2.66028404e-01 -1.08941197e-02
4.12379891e-01 2.81471312e-01 -8.13814461e-01 -4.15086597e-02
-1.97213694e-01 5.45960546e-01 4.08725321e-01 4.28195924e-01
-6.71918511e-01 1.01383102e+00 5.70024967e+00 1.11503363e+00
-1.22282040e+00 7.69300759e-02 7.97200739e-01 -3.65853876e-01
-9.40632761e-01 -3.97438854e-01 -7.04366744e-01 4.41614121e-01
2.69727647e-01 -1.99737176e-01 2.20626518e-01 9.29051280e-01
2.84451276e-01 3.87120128e-01 -9.55986381e-01 1.19063461e+00
2.30328087e-02 -1.78756249e+00 6.31227136e-01 2.51665175e-01
1.11889029e+00 3.77961338e-01 1.93987668e-01 2.73545206e-01
3.72092217e-01 -1.03612018e+00 8.86516988e-01 3.66449803e-01
1.12417388e+00 -8.47442031e-01 3.84978980e-01 3.15788686e-01
-1.24366820e+00 2.86152244e-01 -5.84109366e-01 1.56024262e-01
5.11367857e-01 1.03873742e+00 -8.16560328e-01 6.57247245e-01
8.53676558e-01 6.60132408e-01 -2.82018125e-01 9.79968846e-01
-2.41213620e-01 3.05671453e-01 -6.67752683e-01 1.51634037e-01
4.40357894e-01 -1.72928050e-01 6.75557017e-01 5.92802107e-01
8.85954142e-01 2.03599423e-01 -9.87012386e-02 1.39036846e+00
-3.83689165e-01 -2.88233072e-01 -7.94487953e-01 2.26192132e-01
6.79760516e-01 9.99507129e-01 -5.26386380e-01 -3.28035206e-01
-1.33505212e-02 8.39410722e-01 1.25527084e-01 3.25214744e-01
-7.84732878e-01 -2.51652151e-01 9.95104611e-01 5.22997856e-01
7.79117107e-01 -2.71845430e-01 -7.09117770e-01 -8.38535964e-01
1.43573537e-01 -5.74906170e-01 -2.11704001e-01 -1.09487259e+00
-1.65762949e+00 1.06271160e+00 -1.47032186e-01 -1.34924650e+00
3.87713616e-03 -3.90096009e-01 -4.54240829e-01 9.99114215e-01
-1.46407545e+00 -1.49913597e+00 -4.30150986e-01 3.88893843e-01
5.58074415e-01 4.62010317e-02 5.16998708e-01 4.39928859e-01
-9.65669826e-02 5.09468019e-01 -1.91918656e-01 -6.32704049e-02
3.01223218e-01 -1.13330460e+00 1.37629175e+00 5.83127677e-01
-8.98660123e-02 3.64323407e-01 1.29212320e-01 -9.95247364e-01
-1.34407854e+00 -1.48609245e+00 6.45007730e-01 -6.20099604e-01
3.08606505e-01 -3.28398049e-01 -8.36813092e-01 4.30976540e-01
-1.03036270e-01 6.75245821e-02 2.91020554e-02 -3.49128962e-01
-2.96427906e-01 -1.80756226e-01 -1.41048133e+00 7.09970534e-01
1.55706799e+00 -2.97994524e-01 -4.75317597e-01 2.52038032e-01
1.27901125e+00 -8.72335911e-01 -1.13690591e+00 6.71200037e-01
1.91498280e-01 -8.78679335e-01 1.31614947e+00 -1.64080456e-01
7.86353052e-01 -3.38431776e-01 -6.83437213e-02 -1.55316174e+00
-4.89503264e-01 -3.75479579e-01 -9.48769376e-02 9.89621937e-01
3.97284299e-01 -5.92654169e-01 1.13471448e+00 4.31103587e-01
-6.21587336e-01 -1.28909755e+00 -1.17828155e+00 -7.53793836e-01
4.56896782e-01 -6.12992227e-01 1.38796151e+00 7.35790610e-01
-9.18219209e-01 2.09723070e-01 1.62657406e-02 7.82838687e-02
7.73602605e-01 1.45160884e-01 8.64380836e-01 -1.14980984e+00
1.57053545e-01 -6.15454137e-01 -3.05050462e-01 -1.61009002e+00
-7.72842392e-02 -1.09802997e+00 -1.57698721e-01 -2.09006500e+00
-4.03312653e-01 -8.92622292e-01 1.95497170e-01 4.20829743e-01
-1.23495527e-01 4.16332185e-01 1.77682668e-01 2.57862508e-02
2.01251164e-01 1.26681340e+00 1.94093132e+00 -5.02312958e-01
-1.89826697e-01 -1.68901801e-01 -6.59704685e-01 4.41646606e-01
6.03742003e-01 -3.94059330e-01 -5.03145814e-01 -1.07040620e+00
1.79806650e-01 -1.51070356e-01 4.94408607e-01 -1.02264559e+00
4.62697707e-02 1.25228867e-01 5.81288576e-01 -1.20166647e+00
7.03893483e-01 -8.23581457e-01 3.10771048e-01 1.68438762e-01
2.93009251e-01 2.14775950e-01 2.92648852e-01 3.86442631e-01
-1.99928656e-01 2.04744712e-01 6.89442337e-01 -3.03799421e-01
-2.93371946e-01 9.73353744e-01 1.50353506e-01 -5.55325449e-02
8.70347619e-01 -6.48760438e-01 -1.15563385e-01 -2.58239597e-01
-3.56892139e-01 2.21524775e-01 6.13674164e-01 5.09878457e-01
1.05369532e+00 -1.87081087e+00 -7.75366366e-01 5.10563135e-01
-1.39835805e-01 1.04921830e+00 6.26106024e-01 3.57850641e-01
-7.00039566e-01 2.26216033e-01 -5.45799907e-04 -8.67721975e-01
-7.42101550e-01 2.04414457e-01 3.63427997e-01 -2.00022846e-01
-8.82046938e-01 1.14688945e+00 4.24973458e-01 -8.00382316e-01
8.46919697e-03 -7.42738068e-01 4.10240412e-01 -1.30265728e-01
1.00713097e-01 2.20691159e-01 3.68681371e-01 -3.59414101e-01
-8.28321800e-02 7.87126601e-01 1.74217330e-05 -1.42582394e-02
1.68476605e+00 3.93259674e-01 -7.31339231e-02 -1.02708735e-01
1.06820202e+00 -1.42118037e-01 -1.56448460e+00 -7.84180686e-02
-6.94051802e-01 -5.25257409e-01 3.20621401e-01 -7.31358647e-01
-1.61317372e+00 8.25653672e-01 4.02803272e-01 1.66430846e-02
1.02464616e+00 -2.21056165e-03 1.39555979e+00 -5.25208283e-03
5.15919089e-01 -6.39243245e-01 1.10946745e-01 6.99393153e-01
1.31730652e+00 -8.52556109e-01 -9.72785726e-02 -4.19342160e-01
-3.90150666e-01 6.55486166e-01 8.55492651e-01 -2.87949741e-01
8.03399622e-01 3.37526888e-01 1.38743920e-02 -5.01532435e-01
-6.73814833e-01 6.51024058e-02 3.58683795e-01 7.49735892e-01
1.20757103e-01 3.16343941e-02 8.28137994e-02 4.57331926e-01
-5.40578604e-01 9.82670113e-02 2.39515364e-01 6.47987247e-01
-8.95785540e-02 -1.13959467e+00 -4.17079955e-01 6.31399214e-01
2.37981379e-02 -3.13260972e-01 -3.28273863e-01 6.65823996e-01
2.30996192e-01 4.40035343e-01 3.26868117e-01 -5.36101758e-01
8.45999956e-01 -2.69219548e-01 5.68597913e-01 -5.36388397e-01
-6.72536552e-01 8.13291818e-02 -3.61058295e-01 -5.28338194e-01
-7.12967217e-02 -2.77533472e-01 -1.13850069e+00 -5.64446986e-01
-1.25120670e-01 -9.70274061e-02 6.20203018e-01 4.39317465e-01
8.74984264e-01 8.25243831e-01 5.90268373e-01 -1.45972872e+00
-5.68007052e-01 -9.51203883e-01 -2.64976889e-01 2.61654943e-01
1.77847430e-01 -7.53987670e-01 -2.46994793e-01 -2.90375292e-01] | [8.856958389282227, -3.659865617752075] |
fc4a0c2e-2d33-40b0-aab1-84201fba9fc9 | domain-adaptive-transfer-learning-on-visual | 2010.03071 | null | https://arxiv.org/abs/2010.03071v1 | https://arxiv.org/pdf/2010.03071v1.pdf | Domain Adaptive Transfer Learning on Visual Attention Aware Data Augmentation for Fine-grained Visual Categorization | Fine-Grained Visual Categorization (FGVC) is a challenging topic in computer vision. It is a problem characterized by large intra-class differences and subtle inter-class differences. In this paper, we tackle this problem in a weakly supervised manner, where neural network models are getting fed with additional data using a data augmentation technique through a visual attention mechanism. We perform domain adaptive knowledge transfer via fine-tuning on our base network model. We perform our experiment on six challenging and commonly used FGVC datasets, and we show competitive improvement on accuracies by using attention-aware data augmentation techniques with features derived from deep learning model InceptionV3, pre-trained on large scale datasets. Our method outperforms competitor methods on multiple FGVC datasets and showed competitive results on other datasets. Experimental studies show that transfer learning from large scale datasets can be utilized effectively with visual attention based data augmentation, which can obtain state-of-the-art results on several FGVC datasets. We present a comprehensive analysis of our experiments. Our method achieves state-of-the-art results in multiple fine-grained classification datasets including challenging CUB200-2011 bird, Flowers-102, and FGVC-Aircrafts datasets. | ['Vassilis Athitsos', 'Ashiq Imran'] | 2020-10-06 | null | null | null | null | ['fine-grained-visual-categorization'] | ['computer-vision'] | [ 5.06470203e-02 -3.52348804e-01 -2.87627906e-01 -5.32671630e-01
-5.02075076e-01 -7.91516125e-01 8.22493017e-01 -7.25500286e-02
-5.54770947e-01 6.50152028e-01 1.26642823e-01 -1.10710017e-01
7.02356454e-03 -7.25942433e-01 -9.59953845e-01 -3.81962389e-01
1.12391599e-01 4.82097864e-01 3.04834008e-01 -2.45834053e-01
1.47471339e-01 5.62017620e-01 -1.76062298e+00 8.56225967e-01
6.41783476e-01 1.41525578e+00 4.71325498e-03 7.96299040e-01
-2.43661851e-01 9.30441380e-01 -6.18348241e-01 -2.88422644e-01
3.48070860e-01 9.55519229e-02 -1.15435231e+00 2.40203030e-02
1.02816963e+00 -3.28174859e-01 -1.60801813e-01 8.37663352e-01
2.81487525e-01 1.83071324e-03 9.36304033e-01 -1.59509230e+00
-1.33842683e+00 2.56238103e-01 -7.38710761e-01 4.18643326e-01
-3.96041095e-01 2.62580186e-01 8.90576541e-01 -9.97780025e-01
3.31337899e-01 1.41164088e+00 8.62462401e-01 5.92275739e-01
-1.29846144e+00 -8.17765594e-01 5.38212597e-01 4.30311352e-01
-1.43458247e+00 -1.67453095e-01 5.93400896e-01 -7.76377201e-01
9.05723810e-01 9.54747424e-02 1.72759384e-01 1.02263355e+00
-2.31777862e-01 5.67084849e-01 1.10234904e+00 -4.74488884e-01
1.97286438e-02 9.29537192e-02 4.37974602e-01 8.24447453e-01
3.18406373e-02 3.01746219e-01 -1.85188100e-01 6.42807931e-02
6.43193543e-01 1.33051246e-01 -9.92002711e-02 -3.36229116e-01
-1.15324581e+00 1.11814272e+00 1.11640728e+00 2.43610516e-01
-1.53164536e-01 1.98196709e-01 6.36255562e-01 4.19201165e-01
6.46655798e-01 4.76402253e-01 -8.50243092e-01 2.65297979e-01
-5.30892730e-01 5.74138016e-02 4.25730795e-01 9.14056897e-01
7.32118309e-01 1.57829300e-01 -7.17059910e-01 1.10784626e+00
7.19801039e-02 3.55459273e-01 7.36112237e-01 -7.56937802e-01
4.05148417e-01 6.85926080e-01 -5.92920035e-02 -8.62899005e-01
-3.51181775e-01 -6.07716024e-01 -1.14887583e+00 3.81894976e-01
3.75867754e-01 4.28411849e-02 -1.35040236e+00 1.72882032e+00
2.71077037e-01 1.56820938e-01 -1.17077783e-01 9.25023079e-01
1.42312813e+00 5.49965322e-01 3.53370517e-01 2.53303707e-01
1.21642435e+00 -1.65211165e+00 -3.91087711e-01 -2.86386997e-01
4.42236811e-01 -4.41540569e-01 1.57253993e+00 3.13416779e-01
-4.19883519e-01 -1.21789300e+00 -9.45635140e-01 -2.01170638e-01
-7.81044662e-01 4.09130275e-01 6.63715065e-01 4.24698442e-01
-9.78786409e-01 5.32873690e-01 -3.69540215e-01 -3.74078423e-01
9.96808052e-01 2.91721523e-01 -6.32445812e-01 -2.07895249e-01
-7.25756466e-01 5.01579165e-01 5.14460325e-01 -1.05540447e-01
-1.13125694e+00 -9.99243557e-01 -6.58870995e-01 1.36624947e-01
2.88351774e-01 -5.90750515e-01 1.27112484e+00 -1.12327754e+00
-1.05829751e+00 1.12248099e+00 2.74303883e-01 -5.56962311e-01
3.44447941e-01 -1.12774335e-01 -3.44646275e-01 9.54050571e-02
2.66708255e-01 1.08073545e+00 1.11515677e+00 -1.30676258e+00
-7.12408483e-01 -3.92636061e-01 1.82450846e-01 -2.19085678e-01
-5.80393612e-01 -1.70624390e-01 -4.58503634e-01 -8.45578790e-01
-6.07689619e-01 -8.26780081e-01 -4.56755832e-02 2.26817727e-01
-2.33153537e-01 -5.71241617e-01 1.04395688e+00 -4.01344776e-01
8.90159845e-01 -2.13616538e+00 -1.34061679e-01 -2.91929007e-01
3.86379510e-01 7.51765549e-01 -5.75145721e-01 4.88481633e-02
-5.90838045e-02 3.08545530e-01 -8.69577285e-03 -1.15500882e-01
-3.65623832e-02 3.40726413e-02 -4.34028864e-01 2.87579626e-01
5.72165787e-01 1.33950162e+00 -4.97727603e-01 -3.00859869e-01
2.38824978e-01 2.29740158e-01 -6.28940284e-01 4.70363528e-01
-2.98175573e-01 2.11436957e-01 -2.28259474e-01 8.03074062e-01
7.55469382e-01 -6.90793395e-01 -2.79561222e-01 -6.49490416e-01
1.29609108e-01 -6.31601512e-01 -6.23582661e-01 1.53539002e+00
-4.58679795e-01 5.47788084e-01 -1.59748167e-01 -1.12856114e+00
8.33849907e-01 -2.27677941e-01 -2.51954436e-01 -8.01739872e-01
2.99287647e-01 -2.30339810e-01 1.14646457e-01 -2.42793903e-01
2.88627088e-01 -5.35930283e-02 -1.21702738e-01 2.33662233e-01
5.11878133e-01 -1.06195331e-01 2.88039774e-01 1.17599882e-01
7.18700707e-01 -8.48982409e-02 3.49925905e-01 -4.32936043e-01
5.18136680e-01 2.38670185e-01 2.38394096e-01 9.20612752e-01
-5.23251355e-01 6.57802641e-01 1.43806487e-01 -8.31746936e-01
-8.65123868e-01 -7.30386078e-01 -7.98018500e-02 1.68720162e+00
3.70134823e-02 -3.64413410e-01 -5.69251716e-01 -1.30484653e+00
4.51241285e-01 3.06766361e-01 -1.52175152e+00 -3.33877623e-01
-8.81993771e-02 -7.48234928e-01 6.14401221e-01 1.19572783e+00
8.22723508e-01 -1.21252143e+00 -1.39502093e-01 -6.32324815e-02
5.67969196e-02 -1.31120396e+00 -4.10473615e-01 3.18096817e-01
-6.00069106e-01 -1.15867829e+00 -6.47261679e-01 -9.10787046e-01
5.57261884e-01 4.85384554e-01 1.29458535e+00 1.57109052e-01
-6.53720617e-01 2.98647165e-01 -4.13899690e-01 -5.98867238e-01
-1.37152046e-01 1.94138736e-01 -3.16065550e-02 1.18014082e-01
2.99404353e-01 -5.80282919e-02 -4.57102776e-01 4.80417520e-01
-6.47874594e-01 -3.90269198e-02 5.20064771e-01 1.33932447e+00
7.65780151e-01 -2.77020216e-01 7.45823383e-01 -9.92406130e-01
4.56597358e-01 -3.24261904e-01 -7.10586548e-01 3.01094025e-01
-4.41811293e-01 9.32810009e-02 1.00558794e+00 -5.51866055e-01
-9.78050411e-01 -1.65370464e-01 -1.68672241e-02 -9.37262416e-01
-5.22114038e-01 1.59471855e-01 2.84766611e-02 -3.93590540e-01
1.01336086e+00 -3.94061953e-03 -2.63257623e-01 -6.97172940e-01
7.64560163e-01 8.54045510e-01 7.00744987e-01 -5.02676189e-01
7.62331545e-01 5.01402378e-01 -2.40607142e-01 -6.40277207e-01
-1.21742868e+00 -3.90215904e-01 -9.78739560e-01 1.15874134e-01
9.31777775e-01 -1.10254157e+00 -6.38380408e-01 5.95224679e-01
-8.40099931e-01 -6.42212033e-01 -3.88215303e-01 1.05205681e-02
-2.58757621e-01 -6.93251416e-02 -6.09061897e-01 -1.92572638e-01
-4.93074417e-01 -8.77724290e-01 1.12577653e+00 2.01820254e-01
1.55979753e-01 -9.71567631e-01 1.20312357e-02 4.92617995e-01
6.29827023e-01 7.70559013e-02 9.10605669e-01 -6.64130270e-01
-2.81166524e-01 1.25084259e-02 -9.39525425e-01 6.47945344e-01
3.01290184e-01 1.88933257e-02 -1.27737737e+00 -5.90161204e-01
-7.76903927e-01 -1.10583472e+00 1.35082543e+00 1.97856858e-01
1.96839988e+00 -2.21380964e-01 -3.82427901e-01 9.27187920e-01
1.39898551e+00 -8.60012323e-02 1.85595021e-01 3.63987714e-01
1.08223319e+00 4.43061233e-01 7.84629285e-01 2.20365867e-01
3.15254480e-01 5.75989604e-01 6.85422421e-01 -4.03670639e-01
-5.91897607e-01 -1.60059646e-01 -3.39119941e-01 2.80548781e-01
-1.04568653e-01 -2.63818264e-01 -8.66343617e-01 6.87107921e-01
-1.73277855e+00 -6.49309397e-01 6.54053390e-02 1.75282717e+00
6.04900360e-01 -5.88458031e-02 1.44090518e-01 4.60796915e-02
7.05009937e-01 7.38462359e-02 -7.99009085e-01 -4.04059261e-01
-6.67063296e-02 2.14887515e-01 3.82123441e-01 1.23457104e-01
-1.68804681e+00 1.38829613e+00 6.26881361e+00 9.45446372e-01
-1.34725845e+00 2.66919553e-01 9.46739078e-01 5.23118861e-02
3.10820162e-01 -7.42969155e-01 -8.31423640e-01 2.85055965e-01
6.87258244e-01 1.01228312e-01 2.91752934e-01 1.07991290e+00
-5.12376130e-01 3.72720033e-01 -1.03706682e+00 1.14851391e+00
2.21191391e-01 -1.69754362e+00 3.25103939e-01 -1.46866441e-01
9.66827035e-01 3.01051021e-01 1.60852671e-01 7.53129065e-01
5.07496178e-01 -1.33179414e+00 5.32607257e-01 8.29917863e-02
1.39787734e+00 -7.04150438e-01 7.41389811e-01 1.70807928e-01
-1.30885065e+00 -3.15894067e-01 -6.23597682e-01 -2.79650092e-02
-4.61601913e-01 2.18224764e-01 -8.11354935e-01 3.64739805e-01
1.36688983e+00 8.62522602e-01 -9.93599534e-01 8.50676119e-01
-9.44068357e-02 7.36126781e-01 1.22659199e-01 3.39519531e-02
3.35277468e-01 3.69575113e-01 -1.85418233e-01 1.29526472e+00
-2.35210657e-01 -7.95827731e-02 2.95546472e-01 7.69362748e-01
-5.98376870e-01 -2.99182590e-02 -4.82654452e-01 -2.08251357e-01
1.64099112e-01 1.58218372e+00 -6.59473419e-01 -5.87490261e-01
-4.76179391e-01 9.94349539e-01 8.96053493e-01 2.14003354e-01
-8.83576095e-01 -4.45493519e-01 6.95611298e-01 -1.38866141e-01
7.46655226e-01 2.69296527e-01 -8.85104910e-02 -1.17774689e+00
-3.74170691e-01 -9.56559777e-01 6.54607475e-01 -7.81220496e-01
-1.87824881e+00 1.01761675e+00 -1.88324854e-01 -1.10569608e+00
-1.88304484e-02 -9.75956619e-01 -4.56651658e-01 6.41704738e-01
-1.70959222e+00 -1.68100595e+00 -8.70769143e-01 9.21501756e-01
7.14325488e-01 -5.57747066e-01 8.67851377e-01 2.20112264e-01
-4.46025312e-01 9.33175206e-01 1.07915670e-01 3.13095301e-01
9.51464176e-01 -1.32957935e+00 6.08641446e-01 5.80969810e-01
2.43730158e-01 1.56520918e-01 2.04363186e-02 -4.66980517e-01
-9.56379950e-01 -1.81710792e+00 2.73114353e-01 -4.49498862e-01
5.71112514e-01 -4.96528625e-01 -1.03361082e+00 7.37527847e-01
2.04759881e-01 9.14555073e-01 6.93076074e-01 1.89957634e-01
-8.30168247e-01 -2.25915819e-01 -1.26796091e+00 3.83435227e-02
1.26377285e+00 -4.82759714e-01 -7.17695832e-01 3.11881483e-01
1.05961072e+00 -1.02431402e-01 -6.47881866e-01 7.05981731e-01
4.92652237e-01 -6.78173304e-01 1.11862445e+00 -1.34558165e+00
4.82125819e-01 -3.81590724e-01 -3.93948406e-01 -1.62339056e+00
-8.98418009e-01 1.33251294e-01 -3.30346972e-02 1.26369274e+00
1.08146429e-01 -4.30320591e-01 6.59244478e-01 -4.25196299e-03
-1.00799292e-01 -4.95019346e-01 -4.30662125e-01 -9.07750130e-01
4.53064203e-01 -2.48820037e-01 6.29861712e-01 1.04575551e+00
-7.42022276e-01 5.43322861e-01 -3.53209913e-01 3.36580761e-02
5.55553019e-01 5.29802382e-01 9.37809646e-01 -1.43903220e+00
-2.29484051e-01 -2.74410546e-01 -3.97514552e-01 -6.84045374e-01
3.53073627e-01 -1.00214612e+00 -7.52792507e-02 -1.50299346e+00
4.42621112e-01 -3.69050175e-01 -5.16250730e-01 9.17913973e-01
-2.04985753e-01 1.00487936e+00 4.43259865e-01 5.38442954e-02
-7.84331024e-01 7.93638468e-01 1.30648291e+00 -5.82722843e-01
1.67318135e-01 -2.95936048e-01 -1.00432849e+00 7.88628876e-01
7.06744552e-01 -1.51830986e-01 -3.92795086e-01 -5.00041306e-01
-4.93198633e-01 -5.90414047e-01 5.24696410e-01 -9.80453074e-01
-8.40445906e-02 -1.05457902e-01 9.86705542e-01 -5.87299228e-01
2.49602705e-01 -6.20308220e-01 -3.38926136e-01 4.76531386e-01
-3.66878271e-01 -1.18737407e-01 7.71929324e-01 5.81932724e-01
-3.65218103e-01 2.44329765e-01 1.07585168e+00 -8.44077300e-03
-1.28653896e+00 6.46888435e-01 2.08354518e-02 3.76309514e-01
1.13686621e+00 2.16164127e-01 -7.91363835e-01 1.86473355e-02
-6.33776009e-01 2.32374817e-01 1.67410508e-01 8.09977412e-01
4.63848472e-01 -1.59589148e+00 -8.76217961e-01 4.35678542e-01
7.65283108e-01 -2.40327999e-01 4.10086006e-01 2.11181745e-01
-2.64652669e-01 6.50629103e-01 -7.31040001e-01 -8.43467534e-01
-1.47249520e+00 1.14164877e+00 2.36115128e-01 -3.04100603e-01
-1.77803978e-01 1.16334951e+00 8.83967519e-01 -8.16297114e-01
2.22216129e-01 -4.52373743e-01 -4.62496042e-01 1.04486763e-01
8.00625443e-01 1.47375762e-01 3.07463348e-01 -6.16400301e-01
-4.61367309e-01 7.62744606e-01 -3.76327842e-01 5.12117624e-01
1.31099081e+00 2.17463002e-02 3.16438138e-01 1.11845888e-01
1.50349736e+00 -3.52392823e-01 -1.45449901e+00 -4.73285288e-01
-4.59484041e-01 -5.99218249e-01 1.63084581e-01 -1.34549224e+00
-1.38924575e+00 1.21501827e+00 1.01826382e+00 5.43501042e-02
1.20960653e+00 2.44196832e-01 2.70985305e-01 5.45025110e-01
2.57248163e-01 -7.32964396e-01 3.19132566e-01 6.77964866e-01
1.32354236e+00 -1.82513297e+00 -2.96338648e-01 -4.05039042e-01
-6.98210239e-01 8.91439199e-01 1.23036027e+00 -2.41716892e-01
7.97286689e-01 4.65117991e-02 1.81061864e-01 2.87208008e-03
-8.45503092e-01 -4.45088774e-01 7.70536363e-01 7.83297122e-01
2.04374403e-01 1.08959921e-01 2.34388098e-01 9.32339370e-01
2.26936005e-02 1.12332612e-01 8.79598409e-02 6.83941483e-01
-3.33738834e-01 -9.19705510e-01 -2.52556741e-01 6.73525989e-01
-4.30516541e-01 -2.91314334e-01 -5.99577248e-01 1.04769003e+00
2.80136883e-01 8.40819955e-01 2.99003631e-01 -5.51734805e-01
4.39258993e-01 -1.62073478e-01 5.88950336e-01 -5.43726444e-01
-7.24268794e-01 -3.21621567e-01 -9.50631406e-03 -7.55304039e-01
-3.05415958e-01 -1.46160707e-01 -8.52587044e-01 -2.46691287e-01
-1.71078250e-01 -5.02522253e-02 3.19876224e-01 7.85458148e-01
6.30769730e-01 8.78380656e-01 5.34314156e-01 -9.70375836e-01
-5.07494211e-01 -1.07122636e+00 -4.52821046e-01 6.63774610e-01
5.35616040e-01 -9.47479606e-01 -2.04135820e-01 1.39256537e-01] | [9.606483459472656, 2.055370807647705] |
82fb2a4f-acb3-4622-932a-60c36ec3ebf8 | latentgan-autoencoder-learning-disentangled | 2204.02010 | null | https://arxiv.org/abs/2204.02010v1 | https://arxiv.org/pdf/2204.02010v1.pdf | LatentGAN Autoencoder: Learning Disentangled Latent Distribution | In autoencoder, the encoder generally approximates the latent distribution over the dataset, and the decoder generates samples using this learned latent distribution. There is very little control over the latent vector as using the random latent vector for generation will lead to trivial outputs. This work tries to address this issue by using the LatentGAN generator to directly learn to approximate the latent distribution of the autoencoder and show meaningful results on MNIST, 3D Chair, and CelebA datasets, an additional information-theoretic constrain is used which successfully learns to control autoencoder latent distribution. With this, our model also achieves an error rate of 2.38 on MNIST unsupervised image classification, which is better as compared to InfoGAN and AAE. | ['Tanay Dixit', 'Animikh Aich', 'Sanket Kalwar'] | 2022-04-05 | null | null | null | null | ['unsupervised-image-classification'] | ['computer-vision'] | [-9.34743136e-02 7.97939718e-01 -3.30531865e-01 -1.22466624e-01
-3.12071741e-01 -4.26430255e-01 1.06759870e+00 -7.95356214e-01
-1.79230973e-01 8.44119847e-01 5.66784084e-01 -8.74123201e-02
3.72042090e-01 -1.08484471e+00 -1.01722491e+00 -1.07572711e+00
2.84983307e-01 6.16287231e-01 -3.88807058e-01 2.07874849e-01
-1.58985555e-01 1.58665210e-01 -1.31080997e+00 3.51752132e-01
3.75689477e-01 6.67003691e-01 2.34864861e-01 8.04067612e-01
5.82412332e-02 1.32423437e+00 -9.07300234e-01 -3.41791838e-01
4.79401827e-01 -9.44002926e-01 -5.21433949e-01 2.79150754e-01
2.53423691e-01 -8.18669558e-01 -8.91611755e-01 1.07107484e+00
3.68237138e-01 -9.90900397e-02 1.09583211e+00 -1.18909132e+00
-8.72349560e-01 1.18650293e+00 1.19152747e-01 -2.04803720e-01
-4.41708788e-02 3.20966780e-01 1.06945217e+00 -2.83600003e-01
7.40418255e-01 1.30318713e+00 2.68852383e-01 7.17337370e-01
-1.48934913e+00 -6.71117663e-01 -3.39196235e-01 -2.02074766e-01
-1.37022877e+00 -4.13516492e-01 8.45120192e-01 -4.42638516e-01
9.21695948e-01 -3.59923303e-01 1.06743586e+00 1.59837556e+00
5.21654248e-01 7.82691717e-01 1.31926095e+00 -7.34179318e-01
5.06839573e-01 4.37924504e-01 -3.69106025e-01 5.24942696e-01
1.37699768e-01 5.34663081e-01 -3.79389524e-01 7.13302940e-02
1.24895203e+00 -3.97129580e-02 2.28250362e-02 -2.26330817e-01
-9.77222443e-01 1.16617465e+00 2.91140914e-01 1.42171890e-01
-4.96574104e-01 7.14474022e-01 3.85699957e-03 5.44859529e-01
1.64499730e-01 5.31468034e-01 -7.84003958e-02 -1.17939517e-01
-9.13849056e-01 2.56754249e-01 9.96592402e-01 1.24584758e+00
8.71533453e-01 7.63830185e-01 -2.42297128e-02 4.32098031e-01
4.40243483e-01 6.42547309e-01 9.47411776e-01 -1.17694283e+00
1.13375239e-01 2.17679039e-01 -2.55244315e-01 -7.93110013e-01
1.53767467e-01 -4.92368311e-01 -9.84252751e-01 4.66286063e-01
3.46635431e-01 -5.01878202e-01 -1.32017410e+00 1.69289827e+00
-3.52212697e-01 2.89247502e-02 5.20090163e-01 6.59944117e-01
5.39141297e-01 9.49475646e-01 -2.03279689e-01 -6.90572560e-02
9.39152956e-01 -8.37682188e-01 -1.03316021e+00 -3.11987489e-01
1.89424425e-01 -4.61723834e-01 8.45809579e-01 3.47145706e-01
-1.02424872e+00 -5.01235247e-01 -1.33262718e+00 5.68318591e-02
-1.81310683e-01 3.35992843e-01 8.42297554e-01 6.18820071e-01
-1.09719062e+00 4.93058383e-01 -9.59299743e-01 -1.80067807e-01
2.34770998e-01 3.61060739e-01 -4.10916090e-01 1.60571352e-01
-1.15299070e+00 7.62995899e-01 1.02304995e+00 -4.00274038e-01
-1.54587042e+00 -4.75658923e-01 -8.63891780e-01 1.18285693e-01
-1.15243636e-01 -7.16938555e-01 1.05280411e+00 -1.02652836e+00
-2.12368536e+00 4.86470282e-01 1.79550976e-01 -9.33354080e-01
4.47091252e-01 9.04355571e-02 -9.96187031e-02 1.02114469e-01
-2.40301222e-01 1.26752722e+00 1.08303678e+00 -1.08894384e+00
-7.04892427e-02 3.03039610e-01 1.84697270e-01 2.44709980e-02
-4.83669162e-01 -7.19093919e-01 -2.85109609e-01 -7.77210116e-01
2.04840317e-01 -1.07668173e+00 -8.23410824e-02 -4.87636060e-01
-4.44475979e-01 -1.13743655e-01 6.53117776e-01 -4.67251539e-01
8.94388616e-01 -2.09678173e+00 1.03908695e-01 2.04356402e-01
1.85206696e-01 -3.21285158e-01 -4.48590256e-02 4.27202851e-01
-1.74621239e-01 1.72752097e-01 2.40781996e-02 -3.14481258e-01
1.30234405e-01 7.01937258e-01 -6.66685998e-01 3.50895524e-01
-3.38427871e-02 9.63729858e-01 -6.50835395e-01 -1.69688940e-01
1.98640823e-01 4.76942092e-01 -9.77136910e-01 3.82981122e-01
-4.92355555e-01 1.62634119e-01 -1.00839838e-01 1.02997825e-01
2.86646008e-01 -3.53608489e-01 4.82438952e-01 -7.19838440e-02
2.13313311e-01 4.02876973e-01 -8.33821237e-01 1.64399815e+00
-3.82035255e-01 1.06660581e+00 -5.57504237e-01 -7.11886883e-01
9.71004248e-01 6.35393858e-01 2.96276987e-01 -3.99780899e-01
2.69730777e-01 -1.23020910e-01 7.28965774e-02 -1.43325105e-01
2.67093599e-01 -4.31706488e-01 -3.24447677e-02 8.42119694e-01
7.99004614e-01 -3.49738300e-01 -5.54761700e-02 2.76760638e-01
9.05164957e-01 1.33821070e-01 1.96630821e-01 -3.71160120e-01
-1.17419057e-01 -1.02432556e-01 3.18607241e-01 9.90910769e-01
4.90089357e-01 7.25510716e-01 8.55403841e-01 -3.82642418e-01
-1.76345801e+00 -1.16241908e+00 2.21720003e-02 1.94076851e-01
-3.86065751e-01 -5.69458783e-01 -1.05984068e+00 -7.28672028e-01
-2.43283898e-01 1.13108206e+00 -7.80717492e-01 -2.57667810e-01
-2.68828541e-01 -6.56482220e-01 6.94260657e-01 5.09104013e-01
6.24137819e-01 -1.04117322e+00 -2.07458362e-01 1.88695148e-01
-8.13353434e-02 -1.17793429e+00 -9.20939445e-02 3.62773001e-01
-9.19356585e-01 -5.91946959e-01 -6.07292891e-01 -4.32378858e-01
9.90569651e-01 -5.79070568e-01 1.04044473e+00 -4.82053757e-01
1.28659099e-01 1.25479773e-01 -2.19633147e-01 -3.70930016e-01
-1.02167964e+00 2.55655020e-01 1.09430000e-01 -2.42384255e-01
3.95590365e-01 -9.27537560e-01 -2.44599119e-01 1.42398970e-02
-1.02537668e+00 4.69518870e-01 8.14553320e-01 1.04865229e+00
4.07637417e-01 5.13802171e-01 2.31946185e-01 -8.05952549e-01
4.98389363e-01 -4.53645259e-01 -6.77877128e-01 -3.31622064e-01
-7.47984707e-01 6.41187131e-01 8.26223314e-01 -7.56956875e-01
-9.28304017e-01 1.97175235e-01 -2.07825363e-01 -8.52985501e-01
-3.11491430e-01 3.24698657e-01 -1.51574343e-01 5.52158713e-01
7.45522141e-01 5.10980010e-01 2.47686014e-01 -3.78873646e-01
4.33559924e-01 5.19410372e-01 4.22122806e-01 -4.19878304e-01
8.39671433e-01 4.89260107e-01 -2.43398294e-01 -7.83518851e-01
-7.66160190e-01 4.14924890e-01 -4.97607470e-01 1.68134794e-01
1.00152946e+00 -1.23044920e+00 -4.97398287e-01 3.48974228e-01
-1.07981050e+00 -8.12407494e-01 -8.47646654e-01 8.31210196e-01
-8.65834951e-01 -2.43568346e-01 -7.31504083e-01 -6.49719954e-01
-4.86795977e-03 -1.13086224e+00 6.04660213e-01 -5.46249524e-02
-3.55072916e-01 -1.14720154e+00 1.14198541e-02 -1.13669612e-01
3.37102562e-01 2.41720751e-01 8.15164149e-01 -4.46580797e-01
-9.54365432e-01 -1.96670175e-01 2.75501341e-01 7.74283469e-01
1.13293745e-01 -1.51223555e-01 -1.13072324e+00 -2.90655017e-01
9.55933556e-02 -4.20169234e-01 8.59416783e-01 3.93686473e-01
1.04909015e+00 -7.97910392e-01 3.22677679e-02 6.93180084e-01
1.41764069e+00 1.08514920e-01 9.88040030e-01 -4.40388098e-02
5.52181125e-01 2.65361369e-01 -2.74082899e-01 4.46614027e-01
2.09052399e-01 3.77797902e-01 2.51876116e-01 2.79321343e-01
-3.52928370e-01 -8.54002476e-01 7.32005596e-01 1.09400725e+00
-6.80196211e-02 -4.70785528e-01 -6.05918884e-01 3.09433341e-01
-1.44728708e+00 -1.18505466e+00 4.04522330e-01 1.70013607e+00
1.08511460e+00 3.20452183e-01 -3.08366120e-01 1.23117276e-01
2.23607019e-01 1.93271652e-01 -3.74754459e-01 -1.33280396e-01
-1.52271733e-01 1.42014042e-01 6.85622036e-01 6.49681032e-01
-7.33687758e-01 1.06532645e+00 8.22187519e+00 9.23590302e-01
-1.20924735e+00 2.77187452e-02 5.55802941e-01 -4.16868739e-02
-5.37367582e-01 8.29944387e-02 -9.88859236e-01 6.75925672e-01
1.21496046e+00 -5.83646297e-02 7.57780850e-01 9.61711049e-01
-6.40108660e-02 2.42541879e-01 -1.14592814e+00 9.17256176e-01
5.23834862e-02 -1.45006049e+00 3.55849057e-01 5.42716265e-01
1.07533300e+00 -1.10928349e-01 2.84690142e-01 3.95675778e-01
9.59847569e-01 -1.23390925e+00 8.85772109e-01 7.18992054e-01
9.29580390e-01 -9.38407004e-01 7.38371432e-01 6.15576506e-01
-3.70392829e-01 -3.89654599e-02 -7.59615719e-01 -2.57619858e-01
-1.41136616e-01 6.70068145e-01 -1.19758260e+00 -1.96246728e-01
3.50536942e-01 5.26103675e-01 -1.28478661e-01 1.78003266e-01
-8.43149006e-01 1.06181014e+00 -3.98494393e-01 2.10857727e-02
2.90428519e-01 -3.76308411e-02 3.96824419e-01 8.72686863e-01
5.18188119e-01 2.03387551e-02 -4.05805279e-03 1.32155526e+00
-3.65168810e-01 -4.26974148e-01 -9.13882494e-01 -6.30201399e-01
2.12395966e-01 8.92198622e-01 -4.82059598e-01 -5.80383182e-01
8.42925385e-02 1.00810182e+00 9.07909572e-02 5.10266304e-01
-7.86071837e-01 -2.96277087e-02 4.23770577e-01 1.74963087e-01
2.83237189e-01 -4.21235472e-01 -1.39661893e-01 -1.19502664e+00
-5.52543938e-01 -8.44693363e-01 -5.57544008e-02 -8.09900165e-01
-1.08385527e+00 5.49372017e-01 2.94228256e-01 -1.08719540e+00
-1.11977363e+00 -6.73065186e-01 -3.95808637e-01 6.98168993e-01
-9.43665147e-01 -1.03241503e+00 -7.59057254e-02 3.06841105e-01
2.78544307e-01 -8.78518879e-01 1.07302082e+00 -1.61167495e-02
-2.08863273e-01 7.16372073e-01 2.12911621e-01 5.27877271e-01
3.55367243e-01 -1.56352663e+00 3.05022627e-01 6.22140408e-01
3.61473531e-01 8.75437438e-01 8.80221307e-01 -5.87257862e-01
-1.38595855e+00 -1.01176345e+00 5.13887823e-01 -5.61328113e-01
3.01271230e-01 -6.32044673e-01 -2.90518224e-01 1.19604957e+00
7.58086741e-01 -2.25416124e-01 6.79854095e-01 -2.53733218e-01
-1.49727687e-01 1.18073136e-01 -9.95189130e-01 6.99353814e-01
7.01715887e-01 -6.21956944e-01 -4.10773754e-01 1.91670224e-01
8.72071266e-01 -4.82383043e-01 -9.45662081e-01 -2.46113315e-02
6.22157097e-01 -8.53561103e-01 7.94713736e-01 -2.58396596e-01
8.91352952e-01 -1.41492724e-01 -2.87061870e-01 -1.70073378e+00
-2.80458659e-01 -5.65581679e-01 -5.94061673e-01 1.12150455e+00
3.56816560e-01 -6.13106132e-01 1.15583515e+00 4.06439394e-01
7.84870088e-02 -4.00008231e-01 -4.26215053e-01 -6.62168682e-01
1.53655335e-01 -3.37043434e-01 8.13005149e-01 8.43710840e-01
-4.63125437e-01 3.36888075e-01 -6.89668953e-01 3.35292034e-02
8.02206516e-01 -3.09627444e-01 9.83462572e-01 -6.80801928e-01
-8.30201328e-01 -1.03731975e-01 -4.93924826e-01 -1.48136747e+00
4.49516773e-01 -1.05595279e+00 4.64180075e-02 -1.19466317e+00
6.09494634e-02 -1.98638737e-01 -1.49407431e-01 5.20046890e-01
4.42422658e-01 3.71601224e-01 1.22566663e-01 2.11508065e-01
-6.63446076e-03 6.75159574e-01 1.35387290e+00 -2.59060830e-01
-6.86797500e-03 -2.50601560e-01 -6.62903547e-01 7.28459299e-01
8.50774348e-01 -6.32030427e-01 -7.38347113e-01 -3.67523193e-01
8.63881111e-02 -7.44683901e-03 3.33742470e-01 -1.16753197e+00
2.75034159e-01 -3.80337983e-02 8.57147038e-01 -6.07170045e-01
5.27433395e-01 -9.12811875e-01 4.55189824e-01 3.93668413e-01
-6.26976609e-01 -2.65219092e-01 -1.11164965e-01 3.72232586e-01
-3.38857651e-01 -3.91775310e-01 7.77250350e-01 -4.54428226e-01
-3.94730121e-01 2.50529587e-01 -8.30557168e-01 -1.64999086e-02
8.07645082e-01 -8.29955637e-02 -2.86200494e-02 -8.69772851e-01
-8.14214885e-01 -2.68103033e-01 4.26966161e-01 9.96952802e-02
6.01872265e-01 -1.72894740e+00 -6.23498321e-01 6.79827511e-01
-2.71913290e-01 3.71180363e-02 -5.80097437e-02 2.46633753e-01
-6.35447919e-01 5.39422393e-01 -4.05058920e-01 -6.39883459e-01
-5.31708300e-01 2.50864148e-01 4.26835507e-01 -2.25377560e-01
-7.56707668e-01 6.39475346e-01 1.05269618e-01 -4.61759150e-01
1.03222281e-01 -1.98829576e-01 5.80505431e-02 2.04042643e-02
4.27053273e-01 5.85673340e-02 -4.32328254e-01 -2.25196168e-01
3.94148499e-01 2.68183500e-02 -7.21532479e-02 -4.93019670e-01
1.27841306e+00 1.43213078e-01 4.84606475e-02 5.34190297e-01
1.41696215e+00 -2.97769066e-02 -1.54802430e+00 -7.74154440e-02
-4.84543741e-01 -2.07112417e-01 1.96519703e-01 -5.33548057e-01
-1.10734224e+00 9.87685442e-01 4.39630926e-01 1.92722932e-01
7.23073423e-01 -8.52867030e-03 5.31904697e-01 5.37311494e-01
1.91066429e-01 -1.04360521e+00 4.66009796e-01 6.84845328e-01
8.88737679e-01 -8.89493346e-01 -3.87272984e-02 2.83868760e-01
-7.81730056e-01 1.09993362e+00 5.39606929e-01 -5.48426390e-01
6.95371985e-01 6.74159527e-01 4.94175358e-03 -4.90875952e-02
-1.01716328e+00 2.56691456e-01 1.18002199e-01 4.32725906e-01
4.16881680e-01 1.26723483e-01 1.90697238e-01 3.23915631e-01
-1.04810655e+00 -3.76049131e-02 7.40179896e-01 5.08691490e-01
-1.77379355e-01 -1.08392704e+00 -2.71835834e-01 4.42372322e-01
-5.41879237e-01 -1.46865144e-01 2.24351715e-02 8.12617004e-01
1.12202600e-01 5.24805188e-01 5.22084475e-01 -4.96027201e-01
-3.29959601e-01 2.69277155e-01 5.49111068e-01 -5.60843349e-01
-1.12944387e-01 2.99539685e-01 -2.10892662e-01 -1.93980590e-01
-1.96947858e-01 -3.86470795e-01 -1.12663877e+00 -4.99056876e-01
-3.10431600e-01 2.61960030e-01 7.88474381e-01 7.60629654e-01
2.00424492e-02 7.73458064e-01 5.45832813e-01 -6.83929205e-01
-6.73802972e-01 -1.21968365e+00 -6.13293469e-01 2.84462124e-01
1.82534769e-01 -3.11950028e-01 -4.98698533e-01 3.40396702e-01] | [11.550790786743164, -0.06229942664504051] |
482179f0-27ae-4d59-9718-bcd06e1bb28a | a-robust-and-low-complexity-deep-learning | 2211.02820 | null | https://arxiv.org/abs/2211.02820v2 | https://arxiv.org/pdf/2211.02820v2.pdf | A Robust and Low Complexity Deep Learning Model for Remote Sensing Image Classification | In this paper, we present a robust and low complexity deep learning model for Remote Sensing Image Classification (RSIC), the task of identifying the scene of a remote sensing image. In particular, we firstly evaluate different low complexity and benchmark deep neural networks: MobileNetV1, MobileNetV2, NASNetMobile, and EfficientNetB0, which present the number of trainable parameters lower than 5 Million (M). After indicating best network architecture, we further improve the network performance by applying attention schemes to multiple feature maps extracted from middle layers of the network. To deal with the issue of increasing the model footprint as using attention schemes, we apply the quantization technique to satisfy the maximum of 20 MB memory occupation. By conducting extensive experiments on the benchmark datasets NWPU-RESISC45, we achieve a robust and low-complexity model, which is very competitive to the state-of-the-art systems and potential for real-life applications on edge devices. | ['Le Hong Trang', 'Truong Nguyen', 'Nghia NVN', 'Lam Pham', 'Cam Le'] | 2022-11-05 | null | null | null | null | ['remote-sensing-image-classification'] | ['miscellaneous'] | [ 2.12328121e-01 -4.68495518e-01 -1.37400061e-01 -4.95508164e-01
-3.30911219e-01 -7.45865107e-02 3.21301997e-01 -1.78910360e-01
-7.98054457e-01 5.19953489e-01 -2.98070788e-01 -7.15073943e-01
-2.42945731e-01 -1.07268655e+00 -5.95535040e-01 -5.20799637e-01
-2.46848300e-01 2.13828823e-03 2.42340431e-01 1.96842253e-02
1.57307416e-01 7.74248004e-01 -1.80886829e+00 2.31194615e-01
5.40987611e-01 1.45992172e+00 2.65624374e-01 8.57525408e-01
2.87507921e-01 1.14974117e+00 -3.89270425e-01 -1.50967270e-01
3.32240373e-01 8.90879780e-02 -8.20889711e-01 -3.74192804e-01
4.70587432e-01 -8.35413933e-01 -4.16764081e-01 1.22962666e+00
9.73877251e-01 -3.25103216e-02 3.08314174e-01 -8.90788496e-01
-5.95320106e-01 5.89947164e-01 -5.02858400e-01 9.02435124e-01
-6.09246969e-01 3.73556204e-02 9.48922753e-01 -9.52684104e-01
3.12781334e-01 1.10597324e+00 8.21593881e-01 4.93373454e-01
-3.52586597e-01 -6.83649182e-01 1.18858412e-01 5.66044629e-01
-1.89836967e+00 -6.11018181e-01 1.33831739e-01 -1.52455732e-01
1.31047726e+00 2.79726297e-01 4.95095253e-01 5.92922270e-01
2.31393397e-01 6.55761123e-01 8.80301595e-01 -4.53377254e-02
3.68677109e-01 -1.64585784e-01 4.08894837e-01 6.23716056e-01
3.08102340e-01 -1.35384321e-01 -2.54639298e-01 3.15034820e-04
8.40178430e-01 1.85980529e-01 -4.24307734e-01 5.00584364e-01
-9.35581148e-01 7.74856389e-01 9.50999379e-01 2.66477048e-01
-6.74969852e-01 7.40666807e-01 5.10548472e-01 1.16167940e-01
5.63395381e-01 -1.11586049e-01 -7.50227749e-01 2.01448604e-01
-8.83166969e-01 -2.16951340e-01 5.08624375e-01 7.55704641e-01
8.25159252e-01 1.98779538e-01 -1.40399173e-01 5.89829922e-01
4.16893274e-01 6.18275106e-01 3.34297717e-01 -6.39556170e-01
3.35926831e-01 4.25657451e-01 -1.06045395e-01 -7.40591705e-01
-4.89710629e-01 -5.92992127e-01 -1.43315184e+00 8.92287716e-02
-2.23968163e-01 -3.43787044e-01 -1.22843528e+00 1.13441002e+00
1.56076953e-01 6.52839780e-01 8.66554081e-02 8.84695053e-01
1.20466578e+00 1.00611377e+00 1.50114700e-01 3.18399727e-01
1.31177652e+00 -1.08595741e+00 -3.37433249e-01 -4.35386002e-01
6.03833735e-01 -2.86980182e-01 9.58018541e-01 1.74837828e-01
-7.07123578e-01 -7.99680948e-01 -1.31329525e+00 -1.72807112e-01
-6.10999942e-01 6.96819603e-01 8.26178491e-01 7.93960035e-01
-1.34464777e+00 7.49479592e-01 -1.08843279e+00 -3.67065042e-01
7.36470580e-01 4.12986636e-01 8.69169757e-02 1.10066216e-02
-1.21790707e+00 6.31405652e-01 4.24241483e-01 5.47657192e-01
-1.08127689e+00 -6.15391016e-01 -6.72269106e-01 3.98805022e-01
-1.66357696e-01 -4.94357616e-01 9.68219161e-01 -8.81871641e-01
-1.39008379e+00 7.63017178e-01 2.42676541e-01 -6.77089274e-01
1.80714577e-01 -4.48769331e-01 -5.48608124e-01 2.12985322e-01
-2.84673184e-01 1.00020218e+00 7.23969400e-01 -7.71507680e-01
-8.85820150e-01 -3.19552809e-01 1.59420803e-01 1.85939446e-01
-4.53194618e-01 2.15828255e-01 -5.74739873e-01 -2.34816849e-01
1.43572226e-01 -8.54303539e-01 -6.20537579e-01 3.33622187e-01
-1.84398517e-01 -8.18878263e-02 1.09152126e+00 -4.65604782e-01
1.04738462e+00 -2.35712171e+00 -5.58756709e-01 2.51410365e-01
2.03641057e-01 7.39124775e-01 -1.55756131e-01 -3.40453714e-01
1.62897661e-01 4.15480494e-01 3.05235200e-02 -8.20721984e-02
-3.28671634e-01 3.26609164e-01 -6.80507272e-02 6.84177697e-01
1.40725538e-01 9.77696180e-01 -5.62273860e-01 -3.71046782e-01
1.74222097e-01 7.86120236e-01 -4.18722898e-01 -4.02157903e-02
1.51032433e-01 -6.00025058e-02 -5.23826957e-01 6.82176590e-01
8.46023798e-01 -6.74653471e-01 2.89495531e-02 -4.02309358e-01
4.03833427e-02 3.42224419e-01 -1.24604619e+00 1.35753262e+00
-5.50736606e-01 8.32475364e-01 1.79996207e-01 -9.31818545e-01
7.77376711e-01 1.48988158e-01 3.09027135e-01 -6.50324941e-01
3.54840368e-01 1.04623057e-01 -1.51030973e-01 -3.82024556e-01
5.91839492e-01 4.10426617e-01 2.78976202e-01 3.02356333e-01
-7.23760054e-02 5.13656616e-01 -1.55002773e-01 -1.73796147e-01
9.35491204e-01 -1.98216915e-01 2.63958454e-01 -6.06287301e-01
3.96248370e-01 -3.05792123e-01 2.83830732e-01 1.15374112e+00
-4.02971208e-01 2.72032619e-01 -7.60813281e-02 -9.09307718e-01
-8.16196918e-01 -8.11352074e-01 -3.96691084e-01 1.16056406e+00
2.09708169e-01 -2.54288793e-01 -8.50900471e-01 -5.54683149e-01
-4.04190332e-01 1.69949517e-01 -4.49344397e-01 1.09833837e-01
-3.77329379e-01 -1.32621861e+00 8.08047473e-01 7.58385599e-01
1.26523221e+00 -1.10977852e+00 -1.22500229e+00 9.04787555e-02
-5.79757290e-03 -1.29828417e+00 -3.11275711e-03 4.79324043e-01
-1.07096088e+00 -7.72197247e-01 -5.14339447e-01 -7.76551068e-01
3.67583781e-01 4.78680879e-01 1.26067543e+00 5.20590425e-01
-1.92340448e-01 -2.37622127e-01 -3.51921916e-01 -3.68894726e-01
1.32271364e-01 3.98523808e-01 -1.08698361e-01 -2.88875550e-01
5.96974015e-01 -3.47019404e-01 -9.60519791e-01 9.48570743e-02
-9.71953034e-01 -7.37937838e-02 6.82696879e-01 6.46267712e-01
5.89060009e-01 4.23248827e-01 4.81140137e-01 -8.35731208e-01
1.43091038e-01 -4.19642687e-01 -6.55348718e-01 2.12092727e-01
-5.45967460e-01 -1.40684441e-01 5.15955508e-01 -1.48934156e-01
-7.40337908e-01 3.15627873e-01 -5.34554303e-01 -2.35190198e-01
-1.43358752e-01 6.79197252e-01 -9.69236940e-02 -3.91990691e-01
8.10171545e-01 2.96712250e-01 -5.07388771e-01 -4.18525606e-01
3.61749202e-01 1.37745047e+00 3.82291794e-01 2.88757533e-02
2.99378157e-01 6.21370494e-01 -2.25784078e-01 -1.10236526e+00
-8.90428245e-01 -4.58287954e-01 -3.54411662e-01 -5.21892030e-03
9.30773616e-01 -1.54511905e+00 -8.47256601e-01 7.78452814e-01
-1.04492116e+00 -5.99808455e-01 1.37476861e-01 5.71864963e-01
-1.39260530e-01 1.18784919e-01 -1.12971139e+00 -6.56527698e-01
-1.18707383e+00 -1.06613326e+00 1.14239800e+00 2.43195802e-01
5.28028429e-01 -6.44299686e-01 -5.17125607e-01 2.10403036e-02
8.03876400e-01 -1.24153063e-01 6.18764400e-01 -4.26958859e-01
-8.21218491e-01 2.20232055e-01 -7.62749195e-01 6.15890980e-01
-9.59995985e-02 -1.34292111e-01 -1.21057010e+00 -3.41794431e-01
1.12182483e-01 -3.59565914e-01 1.19605100e+00 6.23315871e-01
1.64458334e+00 -2.76291996e-01 -2.72996902e-01 1.13611269e+00
1.90258360e+00 1.26023859e-01 1.09521091e+00 3.84713560e-01
8.94503772e-01 -6.10157847e-02 2.22715005e-01 3.91057909e-01
4.25145447e-01 1.49679691e-01 7.24406958e-01 -4.55409020e-01
8.30432922e-02 2.65634626e-01 1.14062794e-01 6.54611349e-01
-1.89404622e-01 -5.02811074e-01 -8.97388101e-01 6.30066276e-01
-1.63743651e+00 -7.87101388e-01 -2.21421048e-01 1.78709757e+00
3.28334659e-01 2.50766933e-01 -5.18752515e-01 2.92449802e-01
6.01405561e-01 3.87442708e-01 -5.61584592e-01 -1.48653552e-01
-5.09270430e-02 3.93551201e-01 1.16252446e+00 2.49036819e-01
-1.45322442e+00 1.09861207e+00 6.76645088e+00 9.63981211e-01
-1.51247406e+00 3.03329229e-01 1.09293771e+00 4.56536748e-02
2.92518020e-01 -4.73566413e-01 -1.15851581e+00 3.27209115e-01
1.34520268e+00 4.53605533e-01 3.61993074e-01 1.04834080e+00
1.38758689e-01 -8.48757476e-03 -7.99696088e-01 1.11857474e+00
-2.04531223e-01 -1.60742867e+00 1.23185910e-01 -1.91426530e-01
7.90130913e-01 1.17233741e+00 -6.09134994e-02 2.92992264e-01
1.93489656e-01 -1.19610548e+00 5.36473572e-01 1.94721907e-01
1.01962137e+00 -7.98880398e-01 1.15728736e+00 3.02814186e-01
-1.39759994e+00 -9.86047760e-02 -8.93967032e-01 -1.37768283e-01
-7.39088655e-02 7.36927986e-01 -3.21451038e-01 3.34753305e-01
1.37344110e+00 8.91046107e-01 -4.58407074e-01 7.86091566e-01
3.01806200e-02 7.52926171e-01 -5.83506763e-01 -2.78412640e-01
5.35191119e-01 1.46513626e-01 -1.19691253e-01 1.16267216e+00
4.81407464e-01 3.60643804e-01 9.98467673e-04 4.88485754e-01
-4.36347127e-01 -1.19742215e-01 -5.57903945e-01 1.86752439e-01
4.37665522e-01 1.45969236e+00 -9.11731660e-01 -6.10800564e-01
-3.79648983e-01 9.89701450e-01 9.28757414e-02 1.49548456e-01
-1.06447005e+00 -3.58645618e-01 6.88632369e-01 -1.53138354e-01
5.61453760e-01 -1.30315930e-01 -2.65495747e-01 -9.42960024e-01
-6.53231144e-02 -7.71777153e-01 2.59824097e-01 -9.19478059e-01
-9.06370103e-01 1.02674437e+00 -4.37074214e-01 -1.00220144e+00
4.15650874e-01 -8.34829032e-01 -4.82122213e-01 7.41618395e-01
-2.08237743e+00 -8.85129452e-01 -6.86197937e-01 5.05189955e-01
4.29068178e-01 -2.55032573e-02 7.32549012e-01 9.19599533e-01
-4.89442945e-01 5.10512412e-01 1.09605990e-01 3.09859753e-01
-8.15822631e-02 -7.71389961e-01 1.08639860e+00 9.18684125e-01
8.82049277e-02 2.30431259e-01 4.71544228e-02 -2.61371791e-01
-1.30117047e+00 -1.80435777e+00 6.61021113e-01 -8.02719444e-02
4.55287069e-01 -3.34733367e-01 -7.72249460e-01 5.96550345e-01
-4.15810384e-02 5.04455090e-01 5.43595254e-01 -3.45590800e-01
-1.53889600e-02 -2.09030375e-01 -1.10230768e+00 3.80296797e-01
1.20146549e+00 -6.80208206e-01 1.12620860e-01 4.32873905e-01
8.06074798e-01 -4.72188056e-01 -7.70838201e-01 5.51434398e-01
2.33167171e-01 -9.12749887e-01 1.02046394e+00 -4.28209156e-01
2.70129949e-01 -3.13131988e-01 -4.28938061e-01 -9.04855072e-01
-5.33468068e-01 -2.35114187e-01 -1.10816658e-01 7.10045755e-01
4.27285492e-01 -4.19710398e-01 1.06981277e+00 -2.26557523e-01
-7.12971166e-02 -7.54464984e-01 -9.10746992e-01 -5.85399508e-01
-3.57489586e-01 -4.39632177e-01 7.36101210e-01 9.24045026e-01
-8.00620019e-01 5.01854002e-01 -5.63447833e-01 6.87576473e-01
4.37851131e-01 -1.13340020e-01 3.78838897e-01 -1.16761768e+00
-5.71819879e-02 -2.15563297e-01 -5.55924892e-01 -1.53993320e+00
-1.55606121e-01 -5.69724143e-01 2.46187687e-01 -1.68348777e+00
4.65748310e-02 -5.56966603e-01 -4.42518145e-01 7.21751213e-01
-1.06629282e-01 7.16527402e-01 1.24307740e-02 2.53038198e-01
-7.13633955e-01 3.91731530e-01 9.15921330e-01 -2.51513183e-01
-9.11542997e-02 -3.93052921e-02 -5.83153069e-01 8.45333219e-01
9.77931798e-01 -5.09116292e-01 -2.59115547e-01 -1.05734861e+00
3.87778521e-01 -3.35595161e-01 4.74249750e-01 -1.41882718e+00
4.36812580e-01 2.19940767e-01 4.07362252e-01 -7.66204834e-01
7.22889677e-02 -9.28082347e-01 5.32050095e-02 7.79222369e-01
7.80302957e-02 1.67409051e-02 3.21341068e-01 2.98054457e-01
-3.76251489e-02 -1.35282442e-01 9.90061224e-01 -4.73469123e-02
-1.29972100e+00 7.06665039e-01 -4.23484951e-01 -3.83457035e-01
7.33570933e-01 -4.15652059e-02 -6.84277773e-01 -1.47532567e-01
-5.03428578e-01 1.49704531e-01 -1.27725780e-01 1.81719035e-01
5.82383037e-01 -1.17706466e+00 -7.89407194e-01 2.35845208e-01
-3.28260772e-02 8.50818306e-02 4.99174714e-01 5.63089430e-01
-1.22941303e+00 3.85632068e-01 -2.86174297e-01 -7.77385294e-01
-8.92451763e-01 1.90854281e-01 8.32144320e-01 -6.53797388e-02
-6.69503391e-01 8.97528350e-01 -2.09060594e-01 -3.12513083e-01
2.85949200e-01 -5.22664130e-01 -3.44039083e-01 -3.06454986e-01
8.47482562e-01 5.12730837e-01 4.60788876e-01 -5.21677673e-01
-6.92107081e-01 4.86502260e-01 5.30149676e-02 3.61195296e-01
1.44400132e+00 -1.24440752e-01 -4.39272784e-02 -1.51005551e-01
1.17814445e+00 -6.91300511e-01 -1.28964245e+00 -2.53202200e-01
-1.96246922e-01 -2.05977745e-02 8.14408243e-01 -4.90317643e-01
-1.60083497e+00 1.00379515e+00 1.45052135e+00 7.24330470e-02
1.40084791e+00 -3.26358736e-01 8.30664873e-01 9.43503439e-01
3.06218952e-01 -1.18134034e+00 -3.31628770e-01 8.50669742e-01
1.84120759e-01 -1.33663154e+00 4.09807600e-02 -2.10329100e-01
-1.45111620e-01 7.68263340e-01 5.93329191e-01 -2.12514251e-02
1.22085953e+00 5.07038057e-01 4.18282986e-01 -4.73028511e-01
-6.13599360e-01 -4.66516167e-01 9.16860551e-02 5.19480705e-01
2.84591436e-01 4.07967437e-03 -1.43873781e-01 2.61852831e-01
4.90863845e-02 2.23346233e-01 2.98174769e-01 9.58589792e-01
-7.42605567e-01 -3.29470158e-01 6.11648075e-02 5.93756974e-01
-8.94440055e-01 -6.57787561e-01 7.36532034e-03 6.33512378e-01
1.23137586e-01 8.62772942e-01 4.47934091e-01 -4.97876793e-01
2.36294307e-02 -5.43235183e-01 -7.42021576e-02 -3.94608200e-01
-7.50430644e-01 -1.37633249e-01 -3.26170288e-02 -5.59356868e-01
-6.96936369e-01 2.16192231e-01 -1.18332243e+00 -6.03782833e-01
-4.71854985e-01 -9.71898213e-02 9.02484655e-01 8.97334754e-01
5.40587544e-01 5.44663310e-01 5.09756386e-01 -9.12540436e-01
-5.91291368e-01 -9.45741892e-01 -6.06902242e-01 -2.93881577e-02
1.99652433e-01 -6.41755015e-02 -2.46201172e-01 -2.71837525e-02] | [9.233260154724121, -0.6052197217941284] |
b23bea62-4ea3-4f85-a039-a9475ca4c6cc | minimum-error-tree-decomposition | 1304.1103 | null | http://arxiv.org/abs/1304.1103v1 | http://arxiv.org/pdf/1304.1103v1.pdf | Minimum Error Tree Decomposition | This paper describes a generalization of previous methods for constructing
tree-structured belief network with hidden variables. The major new feature of
the described method is the ability to produce a tree decomposition even when
there are errors in the correlation data among the input variables. This is an
important extension of existing methods since the correlational coefficients
usually cannot be measured with precision. The technique involves using a
greedy search algorithm that locally minimizes an error function. | ['X. Ying', 'Y. Ma', 'D. Wilkins', 'L. Liu', 'Z. Bian'] | 2013-03-27 | null | null | null | null | ['tree-decomposition'] | ['graphs'] | [-8.60828683e-02 2.73707777e-01 -3.48233104e-01 -6.32780015e-01
-2.77321041e-01 -4.08159159e-02 2.47603089e-01 -5.02191577e-03
-1.15942687e-01 1.21824551e+00 -2.12138116e-01 -3.57951313e-01
-5.00976145e-01 -1.06020725e+00 -2.85450295e-02 -8.85724187e-01
-2.81964332e-01 9.28679705e-01 3.44707459e-01 -2.73764670e-01
3.52571487e-01 5.43096602e-01 -1.42304647e+00 -5.97583763e-02
5.80135763e-01 1.02865589e+00 7.96792656e-02 7.59896755e-01
-6.67549297e-02 1.22083616e+00 -7.44748473e-01 -2.81136602e-01
1.02665432e-01 -5.56740701e-01 -1.05041039e+00 1.55161932e-01
-2.90546030e-01 -8.22872594e-02 1.12580389e-01 1.12298167e+00
3.69538181e-02 1.72710106e-01 6.95374787e-01 -1.32016289e+00
-3.16579551e-01 6.05768144e-01 -2.38724843e-01 2.52353370e-01
4.19879943e-01 -6.07847273e-01 8.25296640e-01 -5.65342367e-01
2.56502539e-01 1.44636798e+00 9.91543591e-01 1.87449127e-01
-1.44396460e+00 -3.08130175e-01 -1.56066433e-01 4.67526108e-01
-1.44942915e+00 -3.16357642e-01 8.50676596e-01 -3.52637202e-01
1.22236180e+00 3.35665166e-01 6.44145787e-01 3.69516492e-01
4.37207073e-01 1.95594236e-01 1.36910856e+00 -9.56809223e-01
5.16397119e-01 3.16237986e-01 5.37058830e-01 1.10483980e+00
4.66758251e-01 2.04263866e-01 -2.57883936e-01 -3.16708148e-01
8.52852225e-01 1.84826758e-02 1.40880123e-01 -4.85614955e-01
-7.60005176e-01 1.09914172e+00 1.88943446e-01 5.66837251e-01
-4.30668861e-01 1.38618583e-02 1.31983101e-01 5.21389008e-01
3.50514740e-01 3.62529963e-01 -5.23010373e-01 -2.18837485e-02
-1.01617491e+00 -8.23410694e-03 1.09425652e+00 6.71981633e-01
1.00373447e+00 3.60565603e-01 6.15688801e-01 6.47514939e-01
4.67297792e-01 2.66644955e-01 3.82061660e-01 -1.05671525e+00
8.91150311e-02 7.33771324e-01 1.29187275e-02 -1.28726244e+00
-4.54983294e-01 -2.51668274e-01 -8.55316520e-01 6.84137523e-01
2.58089542e-01 -2.31803387e-01 -1.01688993e+00 1.29283953e+00
2.28292733e-01 -5.68648517e-01 4.19213355e-01 3.52077186e-01
7.37174213e-01 6.22258723e-01 -3.41482371e-01 -6.95562601e-01
7.51569510e-01 -8.37087214e-01 -1.06527817e+00 1.02101468e-01
2.81716704e-01 -5.35331249e-01 1.70577139e-01 5.18153548e-01
-1.04067016e+00 -5.05380392e-01 -1.04048026e+00 1.20524906e-01
-4.55958962e-01 -3.43268186e-01 1.13488674e+00 7.04045475e-01
-1.26083839e+00 6.09190941e-01 -9.56809640e-01 -2.04165921e-01
-5.25662780e-01 7.48104274e-01 -3.18066925e-01 4.37038153e-01
-1.24126244e+00 1.52265608e+00 8.18678319e-01 4.57221538e-01
-4.10237014e-01 1.25706658e-01 -7.32394457e-01 8.70518684e-02
1.58275902e-01 -3.99469584e-01 1.31004441e+00 -9.62766349e-01
-1.60198021e+00 3.39785069e-01 -4.80140597e-01 -3.40510398e-01
2.03062862e-01 3.25506300e-01 -4.41307902e-01 9.61128920e-02
-6.07918240e-02 3.22493523e-01 7.21427679e-01 -1.42446244e+00
-7.42284477e-01 -3.24835628e-01 -2.80097067e-01 -1.27195403e-01
3.20525914e-02 -1.50475293e-01 -7.61056393e-02 -1.82682544e-01
8.37273598e-01 -5.77912867e-01 -5.40376604e-01 -5.91150045e-01
-2.24132046e-01 -3.43765140e-01 7.27732658e-01 -6.05055094e-01
1.42823756e+00 -1.70331979e+00 2.83974528e-01 7.39894807e-01
2.40975440e-01 -2.90465474e-01 4.48443294e-01 8.31603944e-01
-2.47337356e-01 -1.14017557e-02 -1.39151305e-01 9.01294872e-03
-2.47117192e-01 7.04888642e-01 -3.00402530e-02 1.73853278e-01
-2.92187184e-01 3.42098415e-01 -7.63678730e-01 -8.64224970e-01
2.44976506e-01 1.63781583e-01 -1.34655505e-01 -5.55392653e-02
1.24761894e-01 -2.07888097e-01 -3.62895578e-01 7.48193681e-01
2.53458381e-01 -7.84567446e-02 4.04771417e-01 1.00883104e-01
-2.96585262e-01 3.16746205e-01 -1.41616571e+00 7.99478650e-01
-7.84091502e-02 5.07539868e-01 -3.96027677e-02 -1.32282650e+00
1.36881649e+00 7.53663361e-01 5.64204395e-01 -1.60376564e-01
2.67882586e-01 1.93314359e-01 -2.33348273e-02 -4.20309037e-01
2.73159862e-01 -8.64828348e-01 2.17746884e-01 1.94578186e-01
1.42474979e-01 -2.95035481e-01 3.99391532e-01 -2.86113650e-01
9.54819441e-01 -4.84540373e-01 8.97491455e-01 -4.07349169e-01
4.29732472e-01 2.56124288e-01 8.27534854e-01 7.37427771e-01
3.49760279e-02 1.87716067e-01 7.56517649e-01 -8.62313211e-01
-9.00082588e-01 -6.51371777e-01 -3.15450013e-01 7.51694441e-01
-3.10871899e-01 -9.11154673e-02 -5.31344593e-01 -4.15472329e-01
-1.54179841e-01 5.49184382e-01 -8.62311959e-01 3.37926865e-01
-1.35403231e-01 -8.12591195e-01 2.51053512e-01 6.23269796e-01
4.48740721e-01 -7.84283102e-01 -6.77209437e-01 4.55815881e-01
-1.23922125e-01 -6.56733871e-01 6.28187239e-01 9.60730255e-01
-1.60379469e+00 -9.97147679e-01 -8.19189474e-02 -9.94457006e-01
9.24313366e-01 8.41043144e-02 9.67655540e-01 4.31095034e-01
1.55838206e-01 -8.03156644e-02 -3.86672080e-01 -3.65403712e-01
-4.42568004e-01 -1.15184836e-01 3.95706035e-02 -4.65990633e-01
7.19210505e-01 -5.18442273e-01 1.74529016e-01 2.42322162e-01
-3.61425281e-01 -1.36427460e-02 4.54943538e-01 1.16399431e+00
4.88499224e-01 1.13170052e+00 3.36585611e-01 -7.79655397e-01
6.23525202e-01 -2.66818970e-01 -7.79684365e-01 3.00516009e-01
-1.02061248e+00 3.68863732e-01 5.05401373e-01 -1.18835969e-02
-1.02171254e+00 2.21851215e-01 -6.29276931e-02 2.01006383e-01
-5.65317683e-02 8.70967090e-01 9.52487737e-02 -2.26056114e-01
5.40697932e-01 4.13532704e-02 7.02638179e-02 -4.76379097e-01
-2.26541385e-01 6.12156093e-01 3.72567922e-01 -3.95211697e-01
5.01955807e-01 1.91752374e-01 4.72988874e-01 -7.93004751e-01
-5.13709843e-01 -4.36070502e-01 -1.00639153e+00 -1.36324689e-01
5.85316002e-01 -4.36731815e-01 -6.15974486e-01 2.57583559e-01
-1.12359464e+00 1.01115115e-01 -8.19829479e-02 7.81637788e-01
-5.37912965e-01 3.85488540e-01 -5.00732720e-01 -1.11150408e+00
-1.40922546e-01 -9.22549963e-01 2.90640146e-01 -6.35855785e-03
-4.58043933e-01 -1.48288763e+00 2.37714410e-01 2.13868782e-01
7.64013901e-02 2.34558731e-01 1.05121934e+00 -6.10774398e-01
4.59220856e-02 -6.36726916e-01 2.22950071e-01 5.37937880e-01
2.53566027e-01 3.35433245e-01 -4.56144989e-01 -2.04505295e-01
4.96635616e-01 -2.48606220e-01 6.15541577e-01 4.80749875e-01
5.99666655e-01 -3.05165797e-01 -2.85513431e-01 1.64878860e-01
1.48045504e+00 9.54479456e-01 5.06980240e-01 5.05641699e-01
4.67480384e-02 4.82123852e-01 1.95727333e-01 2.08703890e-01
1.99057639e-01 3.84473681e-01 8.43514651e-02 -1.05334230e-01
3.53654057e-01 -1.25260755e-01 -6.94071278e-02 1.19975150e+00
-6.42808735e-01 1.08892232e-01 -1.18150151e+00 5.38637042e-01
-1.85080421e+00 -1.30541253e+00 -3.47469240e-01 1.89671767e+00
8.07217777e-01 4.04455066e-01 1.34763382e-02 1.00116920e+00
7.44778812e-01 -1.40122816e-01 -1.45066664e-01 -8.42762172e-01
2.53806591e-01 2.90705800e-01 5.43074250e-01 1.16875124e+00
-8.63444507e-01 7.86088824e-01 9.30184650e+00 2.70313084e-01
-8.16748083e-01 -3.57983895e-02 1.19330831e-01 2.98403978e-01
4.35370691e-02 2.14861825e-01 -8.60460043e-01 1.22199006e-01
8.02949071e-01 -3.06201596e-02 1.16368935e-01 1.04108727e+00
1.17252082e-01 -5.84207118e-01 -9.25271749e-01 4.78048354e-01
-1.37265669e-02 -9.52285409e-01 -2.94999599e-01 -1.05685435e-01
6.19491756e-01 -4.84886527e-01 -3.77359778e-01 1.05992727e-01
7.09060311e-01 -1.03695405e+00 2.62760490e-01 4.01775479e-01
2.92955220e-01 -8.55085790e-01 1.02185440e+00 3.55695724e-01
-9.71850097e-01 -1.36795267e-01 -7.16032445e-01 -9.46242213e-01
-3.97928692e-02 8.94362271e-01 -1.07690835e+00 3.86933088e-01
6.30308747e-01 1.64790884e-01 -1.53444141e-01 9.16953802e-01
-4.67722237e-01 5.86247921e-01 -5.74026406e-01 -5.23018479e-01
3.59060109e-01 -2.91919500e-01 1.72189817e-01 9.67736185e-01
2.16389060e-01 3.19452763e-01 1.53841421e-01 4.44000959e-01
7.05104768e-01 -5.66142313e-02 -6.30412102e-01 2.69147784e-01
5.39804101e-01 7.89083242e-01 -1.05059934e+00 -4.58284169e-01
-5.14276266e-01 4.05613512e-01 2.16890782e-01 2.38847792e-01
-2.81687498e-01 -4.38545942e-01 5.76420762e-02 -1.60759166e-01
5.63033104e-01 -3.48541588e-01 -4.76715803e-01 -8.15973163e-01
-1.95984945e-01 -8.68425608e-01 3.75289261e-01 -6.86575711e-01
-9.12224889e-01 8.84148359e-01 4.13900256e-01 -6.19748950e-01
-7.48095870e-01 -6.05365276e-01 -3.54294807e-01 9.14744377e-01
-7.83439815e-01 -9.28550899e-01 1.13171346e-01 6.95023537e-01
3.58948767e-01 -4.17208463e-01 1.25864673e+00 -2.63309836e-01
-4.64475900e-01 1.35131180e-01 3.42341572e-01 9.27087516e-02
-1.03673272e-01 -1.37615097e+00 -1.50474697e-01 8.23520720e-01
-1.35143191e-01 7.49047756e-01 1.12797248e+00 -7.65303254e-01
-8.41572940e-01 -2.50843018e-01 1.36647689e+00 -2.18222201e-01
6.94618404e-01 5.35512120e-02 -7.52692282e-01 9.94336963e-01
1.13306090e-01 -3.57124090e-01 8.47149670e-01 7.89158225e-01
-8.62634033e-02 -1.12089626e-01 -1.16249394e+00 3.84318084e-02
4.33822066e-01 -1.97741330e-01 -1.24015045e+00 1.33445159e-01
5.26623949e-02 -1.06463097e-01 -1.01605189e+00 5.55568159e-01
6.41717315e-01 -1.07110095e+00 6.78409636e-01 -5.64310014e-01
7.62521997e-02 -3.27471614e-01 -1.74898252e-01 -1.22484517e+00
-8.46045494e-01 -3.55822772e-01 -3.69221233e-02 7.41843641e-01
5.85077226e-01 -7.90695190e-01 9.99518573e-01 8.12215745e-01
2.79391468e-01 -6.04775429e-01 -1.09165037e+00 -9.30546045e-01
-1.70231596e-01 -2.85278022e-01 3.39381903e-01 9.19174194e-01
4.59765941e-01 5.13018608e-01 -4.65907067e-01 2.26079579e-02
7.98836291e-01 1.31805182e-01 1.61566153e-01 -1.76346886e+00
-3.44262719e-01 -3.30434203e-01 -8.22224796e-01 -6.90435469e-01
4.96420078e-02 -2.93571919e-01 1.64235517e-01 -1.72771919e+00
1.79202750e-01 -4.35449481e-01 -3.58703077e-01 8.15058053e-01
1.90556198e-01 -1.35639459e-01 -1.63456768e-01 2.37168774e-01
-2.58333772e-01 1.16840586e-01 8.03089857e-01 2.58340836e-01
-2.48255670e-01 3.52011979e-01 -4.09816712e-01 1.10902905e+00
9.71083522e-01 -9.31232512e-01 -5.71167350e-01 -2.06370413e-01
2.36964211e-01 6.20603144e-01 1.54038846e-01 -1.09327567e+00
4.77311552e-01 -3.39195251e-01 5.18451691e-01 -1.05948567e+00
2.34666914e-01 -1.13550878e+00 7.26072013e-01 7.31361330e-01
-1.67019755e-01 5.90176642e-01 -2.83180535e-01 3.17105353e-01
-6.32223189e-01 -9.56310213e-01 8.43205452e-01 -4.67142135e-01
-8.28013062e-01 -4.30015385e-01 -8.16128135e-01 -6.80465877e-01
9.39949989e-01 -6.12884820e-01 9.81045961e-02 -4.33875144e-01
-9.13075626e-01 -2.55413968e-02 2.12287456e-01 -1.91056803e-01
6.73578203e-01 -1.15397418e+00 -2.50435591e-01 2.64027268e-01
-4.16693985e-01 -6.57106042e-02 -2.88870782e-01 5.26244402e-01
-7.92503536e-01 8.41712415e-01 -5.30173361e-01 -2.87443072e-01
-1.51972210e+00 8.10580492e-01 3.81883711e-01 -3.45201582e-01
-4.90731806e-01 8.40732932e-01 -6.84071898e-01 -1.45131826e-01
3.76375109e-01 -1.26583859e-01 -2.54432648e-01 1.05148800e-01
4.78654206e-01 6.67813480e-01 1.06156118e-01 -4.57753211e-01
-3.05235386e-01 4.68182951e-01 1.18999168e-01 -4.03432876e-01
1.45435441e+00 -1.11366294e-01 -6.89256072e-01 8.72071087e-01
9.36960816e-01 -3.18228066e-01 -6.49299145e-01 -1.71010554e-01
1.17585361e-01 -2.91857690e-01 4.66126651e-01 -6.63915217e-01
-9.67577219e-01 6.94106162e-01 4.17352110e-01 8.41885924e-01
1.15906858e+00 -1.71524420e-01 2.78681964e-01 9.06535506e-01
5.50796449e-01 -1.46831322e+00 -4.42305148e-01 6.65543497e-01
6.93296552e-01 -1.22778225e+00 3.98841411e-01 -3.36478084e-01
-3.82268608e-01 1.47910845e+00 3.52244496e-01 5.49108274e-02
8.98480237e-01 4.56081480e-01 3.12731713e-01 -2.80239344e-01
-9.81729031e-01 -1.38075367e-01 -1.54685974e-01 7.01711595e-01
5.73870838e-01 1.15990214e-01 -7.67825305e-01 2.91284304e-02
-4.96550709e-01 5.75767383e-02 4.33144063e-01 1.16927671e+00
-1.02498281e+00 -1.08587301e+00 -8.38198662e-01 2.38430038e-01
-3.72877777e-01 1.74017921e-02 -3.86420995e-01 9.94235277e-01
3.79007757e-02 1.29060209e+00 -1.85621217e-01 -5.01440704e-01
4.65555303e-02 2.68951595e-01 4.05092627e-01 -4.86523211e-01
-2.68792450e-01 1.21935681e-01 1.55599818e-01 -1.18500710e-01
-7.56576180e-01 -7.59179592e-01 -1.22118163e+00 -6.58989072e-01
-5.55981755e-01 8.34499121e-01 7.80064046e-01 8.48186433e-01
-4.47801113e-01 3.44209105e-01 6.25621855e-01 -4.99762118e-01
-4.06257272e-01 -9.37001586e-01 -7.40383387e-01 -1.34905577e-01
2.05094829e-01 -1.03700209e+00 -4.83396977e-01 1.02997206e-01] | [8.123058319091797, 4.448826789855957] |
7ee27425-0ed4-4621-ba26-0b11f1c1fe21 | revisiting-the-centroid-based-method-a-strong-1 | null | null | https://aclanthology.org/W17-4511 | https://aclanthology.org/W17-4511.pdf | Revisiting the Centroid-based Method: A Strong Baseline for Multi-Document Summarization | The centroid-based model for extractive document summarization is a simple and fast baseline that ranks sentences based on their similarity to a centroid vector. In this paper, we apply this ranking to possible summaries instead of sentences and use a simple greedy algorithm to find the best summary. Furthermore, we show possibilities to scale up to larger input document collections by selecting a small number of sentences from each document prior to constructing the summary. Experiments were done on the DUC2004 dataset for multi-document summarization. We observe a higher performance over the original model, on par with more complex state-of-the-art methods. | ['Gholipour Ghal', 'Demian ari'] | 2017-09-01 | null | null | null | ws-2017-9 | ['extractive-document-summarization'] | ['natural-language-processing'] | [ 3.63329947e-01 2.04783127e-01 -2.12916434e-01 -4.30141091e-01
-1.32925761e+00 -9.06946361e-01 9.23746943e-01 9.05243576e-01
-5.06579697e-01 8.34017456e-01 1.10248649e+00 -6.73266277e-02
-1.19399764e-01 -3.25585932e-01 -3.39454412e-01 -3.86991948e-01
-1.21907130e-01 6.59481943e-01 4.48996276e-01 -2.89436817e-01
1.11037290e+00 2.72033095e-01 -1.36084223e+00 5.84932745e-01
9.45447803e-01 2.38911912e-01 2.61766046e-01 1.26896548e+00
-8.55437666e-02 4.98247832e-01 -1.21196914e+00 -2.89196581e-01
-1.84090763e-01 -8.12071145e-01 -1.06750154e+00 2.08560139e-01
9.11469340e-01 -4.34005439e-01 -1.70220777e-01 7.99267948e-01
6.03062272e-01 3.78375441e-01 9.18068767e-01 -6.46069944e-01
-3.15721363e-01 9.04451013e-01 -4.55661714e-01 3.86504143e-01
7.86343634e-01 -4.85672832e-01 1.27449155e+00 -7.53215909e-01
8.52198005e-01 1.21628201e+00 3.52455854e-01 4.20577228e-01
-1.09921002e+00 -9.11433157e-03 2.28409499e-01 4.98670153e-02
-9.47652519e-01 -7.66243696e-01 5.14038861e-01 -5.19387685e-02
1.30451429e+00 8.24193656e-01 3.61558586e-01 7.79723346e-01
2.71695226e-01 1.01406240e+00 4.34528500e-01 -5.83638966e-01
2.59533107e-01 -2.07766518e-01 7.83105910e-01 3.69547099e-01
5.53318501e-01 -8.46981227e-01 -4.38423157e-01 -6.89865470e-01
-1.78988315e-02 -3.20046127e-01 -4.13640648e-01 6.31073490e-02
-1.48460388e+00 9.17980850e-01 9.94708687e-02 4.76875305e-01
-5.71372986e-01 9.89007652e-02 8.71622741e-01 1.86842948e-01
6.32534146e-01 9.57826912e-01 -1.70140833e-01 -3.49309266e-01
-1.59720886e+00 6.62600636e-01 1.10030353e+00 8.38139117e-01
3.33875835e-01 -2.09245384e-01 -7.69547105e-01 9.35288787e-01
-1.83993086e-01 2.42484957e-01 6.34769619e-01 -1.23418617e+00
7.88031220e-01 2.33284131e-01 2.06847087e-01 -9.64020371e-01
-3.71094853e-01 -3.59390378e-01 -8.45242202e-01 -4.03790981e-01
-1.15443356e-01 -2.14842871e-01 -8.11762512e-01 1.16431201e+00
-2.03638688e-01 -2.51304477e-01 2.95660853e-01 3.86979312e-01
9.36569452e-01 9.84837711e-01 -5.02930164e-01 -6.96536839e-01
9.93079424e-01 -1.25115573e+00 -6.79035604e-01 -7.69559368e-02
5.87953866e-01 -8.19412708e-01 7.19654381e-01 4.97784048e-01
-1.29676282e+00 -2.86524951e-01 -1.17940533e+00 -1.73731461e-01
-7.42913410e-03 4.17567194e-01 5.62017798e-01 3.69436413e-01
-1.42766798e+00 1.01390243e+00 -7.84205973e-01 -7.45316684e-01
7.54680559e-02 2.38256037e-01 -2.70340264e-01 2.15427786e-01
-6.53331876e-01 6.94911063e-01 7.90351987e-01 -3.05039048e-01
-3.74494165e-01 -2.95141667e-01 -7.74939656e-01 2.52996564e-01
1.96683168e-01 -1.01686513e+00 1.55867696e+00 -3.85403097e-01
-1.54105675e+00 3.87965292e-01 -5.96734166e-01 -7.17624009e-01
2.13670641e-01 -4.23866689e-01 1.20220184e-01 5.57672262e-01
2.10469589e-01 5.90473175e-01 4.65348750e-01 -1.26124501e+00
-7.17897058e-01 -1.37678027e-01 -5.35262786e-02 4.45213914e-01
-3.59847307e-01 2.55941898e-01 -4.17713225e-01 -6.39735878e-01
-1.83831695e-02 -8.36459816e-01 -3.34384859e-01 -1.03976858e+00
-9.41445470e-01 -5.10780215e-01 5.96348941e-01 -6.49429798e-01
1.72639799e+00 -1.61156619e+00 4.03560519e-01 -1.08161077e-01
1.16439804e-01 2.70119786e-01 -3.86953980e-01 1.07963777e+00
1.73141032e-01 2.53693670e-01 -2.55875915e-01 -6.38033986e-01
-1.81442592e-02 -1.90940410e-01 -5.26148140e-01 2.23243624e-01
-1.47916600e-01 6.39269590e-01 -1.05969286e+00 -6.43895686e-01
-9.95150506e-02 -1.59934506e-01 -5.31588435e-01 -4.50305343e-02
-2.24931508e-01 -1.80087924e-01 -4.53374088e-01 1.91674218e-01
3.31140906e-01 6.39301091e-02 1.28085032e-01 -4.75937203e-02
-2.05403924e-01 7.16982186e-01 -8.50959718e-01 1.89734375e+00
-2.47063056e-01 9.42414522e-01 -2.95066118e-01 -8.06496382e-01
7.49181390e-01 2.41304398e-01 2.75942981e-01 3.57591696e-02
2.69744508e-02 1.77582458e-01 -9.66975614e-02 -2.18745366e-01
1.48380828e+00 3.26195419e-01 -3.38871717e-01 7.47061968e-01
8.96981955e-02 -5.62862694e-01 1.06322348e+00 8.80209088e-01
1.30502689e+00 -4.32015926e-01 5.92227638e-01 -3.97541493e-01
4.86489743e-01 1.22801311e-01 -6.08008020e-02 1.06692243e+00
3.04962307e-01 1.08431637e+00 7.29014337e-01 1.34538069e-01
-1.06199932e+00 -8.53476882e-01 1.41593382e-01 9.79350030e-01
-1.93454459e-01 -1.18302453e+00 -8.33145380e-01 -7.70262241e-01
-2.13859528e-01 1.22564423e+00 -4.58635867e-01 1.40066892e-02
-8.07047904e-01 -6.95172369e-01 5.59093356e-01 4.27107990e-01
2.56099313e-01 -9.29978371e-01 -5.77672303e-01 3.12662274e-01
-4.41497326e-01 -7.60503888e-01 -8.99479449e-01 -1.99088641e-02
-1.07796979e+00 -6.96983218e-01 -9.01601613e-01 -7.78431594e-01
6.41256452e-01 4.17104632e-01 1.10066485e+00 6.67847842e-02
2.22003087e-01 3.44971865e-01 -5.29850185e-01 -3.61953408e-01
-7.53883958e-01 6.94374263e-01 -1.12334736e-01 -5.93101621e-01
-1.21720403e-01 -2.48227194e-01 -4.11541671e-01 -3.85500342e-01
-9.67913151e-01 -1.27186432e-01 5.77832639e-01 7.14944363e-01
2.64460742e-01 -8.07872042e-02 7.40924895e-01 -9.93079245e-01
1.50449216e+00 -2.07883567e-01 3.90870646e-02 4.38201547e-01
-3.48695397e-01 2.63521850e-01 7.90746927e-01 -1.24926917e-01
-8.72322321e-01 -1.96058244e-01 -2.09268332e-01 1.51162893e-01
-7.08664244e-04 7.16790557e-01 2.43271902e-01 5.99164069e-01
5.81992626e-01 4.81283039e-01 -2.78021723e-01 -4.04852360e-01
5.51359951e-01 9.16557074e-01 7.13810027e-01 -3.02091628e-01
5.50121427e-01 9.23855156e-02 -1.04424797e-01 -1.20604599e+00
-8.93531919e-01 -8.06613982e-01 -8.02777886e-01 1.12800710e-01
5.35854876e-01 -5.33631980e-01 -1.73338488e-01 1.54349446e-01
-1.60288715e+00 9.91654024e-02 -3.23803246e-01 4.81949896e-01
-4.90472108e-01 8.88722777e-01 -6.09606206e-01 -5.18881917e-01
-9.78870034e-01 -6.39882803e-01 1.34770787e+00 2.38668561e-01
-7.32776403e-01 -8.55061591e-01 5.41565180e-01 6.93694800e-02
2.96742082e-01 1.26261383e-01 7.52753973e-01 -1.09424293e+00
-6.93502650e-02 -5.08995533e-01 6.21511079e-02 3.79900455e-01
2.98257917e-01 3.81276280e-01 -5.13485968e-01 -3.79444063e-01
-8.65786821e-02 -6.13718368e-02 1.56558907e+00 5.62443316e-01
8.78655195e-01 -6.45437837e-01 -4.58749533e-01 5.00739664e-02
1.10569453e+00 4.04604152e-02 5.81038356e-01 3.08017313e-01
5.25043070e-01 4.08044428e-01 7.00785100e-01 5.08068144e-01
4.45768446e-01 3.61890674e-01 -1.25204593e-01 3.43629926e-01
-1.68157309e-01 -1.81458846e-01 3.61038834e-01 1.31489980e+00
4.59296964e-02 -8.89588118e-01 -6.58863306e-01 7.79621303e-01
-1.96875656e+00 -1.28109133e+00 -3.47664922e-01 2.16613603e+00
7.81163156e-01 2.22313344e-01 3.02998245e-01 8.20574388e-02
6.30659282e-01 6.72610819e-01 -1.18489079e-02 -8.73283505e-01
-1.76008120e-01 3.99696790e-02 3.74756098e-01 6.44126356e-01
-1.12577319e+00 8.49702239e-01 7.46291065e+00 8.51537824e-01
-8.80314171e-01 -2.98801482e-01 3.48166883e-01 -3.78386676e-01
-4.94123429e-01 -3.60897519e-02 -8.42483997e-01 3.11340570e-01
1.16983700e+00 -7.60771275e-01 3.86152901e-02 5.74196517e-01
2.47147635e-01 -5.16065061e-01 -1.14317727e+00 6.56262577e-01
6.63614035e-01 -1.56507313e+00 4.58569467e-01 -1.52843043e-01
9.88448322e-01 2.60891803e-02 -4.86416698e-01 1.03470683e-01
1.32123277e-01 -5.47572017e-01 5.72892785e-01 4.41612780e-01
3.75561982e-01 -8.00163329e-01 7.83524990e-01 4.60521460e-01
-9.11803544e-01 1.62989125e-01 -5.57402968e-01 9.27314013e-02
2.96641588e-01 5.22543609e-01 -9.83846009e-01 8.89638484e-01
1.41523361e-01 7.59639800e-01 -9.80865180e-01 1.28760290e+00
-1.13694772e-01 7.46963620e-01 -2.20332026e-01 -6.05685651e-01
3.85119230e-01 -7.48811737e-02 1.01093066e+00 1.79207098e+00
4.26400840e-01 1.85467955e-02 1.56036407e-01 1.29981443e-01
-1.76937759e-01 2.99610198e-01 -5.45795739e-01 -8.61468911e-02
5.19102156e-01 1.20816600e+00 -8.61439764e-01 -9.32791591e-01
3.17354128e-02 1.20641029e+00 2.82305211e-01 2.00154498e-01
-4.27744329e-01 -1.05477905e+00 2.25780923e-02 -1.94607809e-01
5.01732767e-01 -3.88067991e-01 -2.52678692e-01 -1.22943795e+00
4.16915305e-02 -7.52848566e-01 3.58918488e-01 -5.77577770e-01
-8.84850085e-01 7.57349670e-01 4.80906039e-01 -1.06964290e+00
-7.66117454e-01 1.32079870e-01 -1.15019298e+00 4.76951778e-01
-1.06716728e+00 -5.45020878e-01 9.94991884e-03 -9.09660608e-02
1.08871496e+00 -2.88939625e-02 9.48692441e-01 -2.99399346e-01
-3.78011733e-01 3.53200167e-01 6.34433091e-01 -2.36339420e-01
8.45779002e-01 -1.63509417e+00 7.40690708e-01 9.44029927e-01
2.02175483e-01 8.27584028e-01 1.17588091e+00 -6.22064769e-01
-1.19129384e+00 -8.87623668e-01 1.35977685e+00 -3.90268356e-01
5.44469953e-01 -1.33018583e-01 -7.81901419e-01 5.86723626e-01
8.91174316e-01 -9.60915744e-01 6.25146627e-01 2.69864827e-01
8.89868960e-02 6.88149184e-02 -7.83892572e-01 7.38992870e-01
7.86992788e-01 -1.14523873e-01 -1.19312549e+00 6.57040834e-01
7.54086018e-01 -3.47842097e-01 -5.71077883e-01 8.91040489e-02
4.69002873e-01 -8.59165549e-01 6.15443647e-01 -3.65869433e-01
6.91617727e-01 -2.37809628e-01 -4.27582450e-02 -1.95914662e+00
-4.02274907e-01 -8.41185927e-01 -5.53910621e-02 1.57690454e+00
4.76776659e-01 -2.96992570e-01 5.78381240e-01 1.72230259e-01
-5.04865885e-01 -4.99269903e-01 -6.20394886e-01 -8.14418375e-01
1.73186705e-01 1.76425800e-01 4.20390546e-01 3.11119914e-01
4.27014947e-01 9.20588315e-01 -1.93506047e-01 -2.74710983e-01
5.44002116e-01 3.06577563e-01 9.26668525e-01 -1.05192566e+00
-1.14222363e-01 -8.12354088e-01 -1.99287325e-01 -1.27481771e+00
1.69724464e-01 -8.23566020e-01 3.33478391e-01 -2.31147647e+00
4.88457739e-01 3.86470467e-01 5.93671529e-03 5.29508963e-02
-5.04117191e-01 1.10489093e-02 3.08540046e-01 2.36984178e-01
-1.13889277e+00 5.96098185e-01 9.37807620e-01 -3.09312910e-01
-4.36162323e-01 1.43739969e-01 -1.08195031e+00 5.44477940e-01
8.98921788e-01 -5.06381571e-01 -3.76069576e-01 -3.98747563e-01
-2.29240075e-01 1.29222989e-01 -1.61055624e-01 -8.88368249e-01
5.32592654e-01 8.71284604e-02 1.16205677e-01 -1.26642191e+00
8.54937807e-02 -3.92094627e-02 -2.86643088e-01 3.92000645e-01
-7.64212608e-01 3.13870937e-01 1.50589928e-01 4.96315986e-01
-3.71351331e-01 -7.61599481e-01 3.25792044e-01 -6.08897172e-02
-1.54153258e-01 -2.76555598e-01 -5.88210702e-01 2.52527129e-02
6.83300495e-01 3.66520286e-02 -6.32668376e-01 -5.38757324e-01
-3.15355092e-01 2.24471688e-01 5.03891110e-01 3.26952904e-01
5.89607954e-01 -9.93271351e-01 -1.23530126e+00 -4.83157724e-01
2.09972952e-02 -1.64298743e-01 -5.03486618e-02 5.31891227e-01
-6.95105612e-01 8.46957743e-01 2.03538746e-01 -4.97750491e-01
-1.65808034e+00 3.03476125e-01 -2.05848143e-01 -5.72975993e-01
-6.67004406e-01 4.80347484e-01 -2.67779827e-01 -1.09365568e-01
9.53041092e-02 -5.50218642e-01 -4.23291773e-01 4.59075332e-01
8.36183727e-01 7.72000968e-01 2.96123743e-01 -4.54271317e-01
-4.31684583e-01 4.27647263e-01 -5.22204876e-01 -3.86069477e-01
1.42908955e+00 -1.90172687e-01 -3.38003129e-01 5.03466547e-01
1.34462917e+00 3.59359890e-01 -5.57561159e-01 6.35951310e-02
3.09249908e-01 -3.32036763e-01 -1.19802624e-01 -4.99345154e-01
-3.81340772e-01 3.28791112e-01 -2.82384366e-01 6.23659611e-01
9.12642658e-01 9.53399837e-02 7.16208816e-01 9.93769288e-01
-3.99301909e-02 -1.04566038e+00 2.91028097e-02 6.74252391e-01
1.30585957e+00 -9.15994823e-01 7.78910637e-01 -1.84474051e-01
-6.69494927e-01 1.45154870e+00 -2.98765469e-02 -4.79711294e-01
1.45415843e-01 -1.51321543e-02 -2.25469172e-01 -1.44734606e-01
-1.14511251e+00 8.81320685e-02 4.57240433e-01 2.82285988e-01
6.57090425e-01 -2.07687784e-02 -9.22553420e-01 3.64079893e-01
-6.40456498e-01 -3.27102184e-01 1.11991334e+00 1.03579247e+00
-8.91207635e-01 -1.09282613e+00 -2.72862583e-01 9.55937386e-01
-5.13987422e-01 -1.83062091e-01 -8.15383196e-01 5.99462092e-01
-7.52359331e-01 1.42944849e+00 2.42901724e-02 -1.22225657e-01
5.46343505e-01 -4.45077270e-02 6.22356236e-01 -1.04745924e+00
-7.87523448e-01 2.40202218e-01 6.18910909e-01 -1.29957125e-01
-3.70485514e-01 -1.14662588e+00 -1.16895771e+00 -2.56511986e-01
-4.18402165e-01 5.74221551e-01 6.39272630e-01 8.32938194e-01
4.36148137e-01 4.81299788e-01 7.79044330e-01 -1.21203625e+00
-8.07946503e-01 -1.36521232e+00 -4.41817522e-01 1.12157688e-01
4.82673228e-01 9.32804942e-02 -4.63960171e-01 -1.30192013e-02] | [12.516295433044434, 9.52125072479248] |
27a36bca-f9c2-4ce5-b8af-1f3ea4979b40 | learning-edge-preserved-image-stitching-from | 2012.06194 | null | https://arxiv.org/abs/2012.06194v1 | https://arxiv.org/pdf/2012.06194v1.pdf | Learning Edge-Preserved Image Stitching from Large-Baseline Deep Homography | Image stitching is a classical and crucial technique in computer vision, which aims to generate the image with a wide field of view. The traditional methods heavily depend on the feature detection and require that scene features be dense and evenly distributed in the image, leading to varying ghosting effects and poor robustness. Learning methods usually suffer from fixed view and input size limitations, showing a lack of generalization ability on other real datasets. In this paper, we propose an image stitching learning framework, which consists of a large-baseline deep homography module and an edge-preserved deformation module. First, we propose a large-baseline deep homography module to estimate the accurate projective transformation between the reference image and the target image in different scales of features. After that, an edge-preserved deformation module is designed to learn the deformation rules of image stitching from edge to content, eliminating the ghosting effects as much as possible. In particular, the proposed learning framework can stitch images of arbitrary views and input sizes, thus contribute to a supervised deep image stitching method with excellent generalization capability in other real images. Experimental results demonstrate that our homography module significantly outperforms the existing deep homography methods in the large baseline scenes. In image stitching, our method is superior to the existing learning method and shows competitive performance with state-of-the-art traditional methods. | ['Yao Zhao', 'Kang Liao', 'Chunyu Lin', 'Lang Nie'] | 2020-12-11 | null | null | null | null | ['image-stitching'] | ['computer-vision'] | [ 3.08567435e-01 -3.32497448e-01 5.61286882e-02 8.64999518e-02
-3.50706637e-01 -4.45336133e-01 5.37264884e-01 -7.36120462e-01
-9.68606621e-02 3.28612298e-01 7.36567229e-02 1.47183970e-01
7.99925178e-02 -6.62362576e-01 -8.53325009e-01 -1.10457611e+00
5.50566494e-01 2.34695598e-01 3.12378466e-01 -2.83943176e-01
4.18519586e-01 1.58128068e-01 -1.31486273e+00 -7.25057274e-02
9.83637154e-01 9.01134551e-01 2.76704818e-01 3.15163076e-01
2.82387197e-01 5.18688083e-01 -4.62814331e-01 -2.92678386e-01
5.55289149e-01 -4.03221071e-01 -2.91541725e-01 6.08076811e-01
9.61434841e-01 -7.10949540e-01 -7.83907592e-01 1.24486363e+00
5.76171398e-01 5.82904480e-02 3.84464473e-01 -1.31426823e+00
-7.33656168e-01 2.45262727e-01 -1.13478637e+00 -2.05186397e-01
2.00306326e-01 4.10956830e-01 5.30487061e-01 -7.67075241e-01
6.02730274e-01 1.36937654e+00 6.20447457e-01 3.58470917e-01
-1.20755136e+00 -1.15779257e+00 -1.77295610e-01 -4.20783311e-02
-1.23571479e+00 -1.90749690e-01 1.15944576e+00 -3.64169449e-01
3.02327782e-01 -7.80781591e-03 7.08209336e-01 1.06601512e+00
4.98903960e-01 6.87276423e-01 1.14358008e+00 -2.42388338e-01
-1.97753191e-01 -2.98703700e-01 -2.79566169e-01 7.72876263e-01
4.36794639e-01 6.45674646e-01 -3.63182694e-01 9.67145413e-02
1.43131685e+00 4.99968112e-01 -5.77846646e-01 -6.77920103e-01
-1.61057019e+00 7.45036900e-01 6.71210527e-01 2.09565952e-01
9.27473232e-02 2.95112252e-01 3.91385645e-01 3.24328929e-01
1.12886295e-01 2.76882529e-01 -5.01212329e-02 3.85248989e-01
-7.73083985e-01 1.99377224e-01 3.69659394e-01 1.08531559e+00
9.88901973e-01 3.30132484e-01 6.55024648e-02 6.75756156e-01
7.30775893e-02 6.74933910e-01 6.96432769e-01 -8.95611644e-01
6.59793019e-01 5.60352862e-01 9.63728726e-02 -1.49680042e+00
-4.95685264e-02 -4.74842936e-01 -1.45606422e+00 3.49081486e-01
1.18247226e-01 -1.31457686e-01 -8.41207504e-01 1.40721893e+00
4.51416761e-01 6.37905836e-01 8.83324593e-02 9.43231583e-01
7.02512801e-01 7.31709003e-01 -5.63458800e-01 -3.02129745e-01
1.22057998e+00 -1.21164739e+00 -7.32825994e-01 -3.99674058e-01
3.17399323e-01 -1.20640588e+00 9.60570931e-01 3.96476984e-01
-7.61727035e-01 -9.99029636e-01 -1.21265101e+00 -2.24864587e-01
4.42580208e-02 1.05567455e-01 5.97303748e-01 5.16707122e-01
-8.74487579e-01 4.65893239e-01 -4.88124043e-01 -1.45866377e-02
1.48999110e-01 3.45314622e-01 -5.32155156e-01 -3.26838136e-01
-9.44577456e-01 5.38266301e-01 8.01271200e-01 8.65063891e-02
-6.86066747e-01 -6.72156870e-01 -1.03801405e+00 -4.32484709e-02
4.08189505e-01 -9.90676403e-01 8.34133267e-01 -1.08991337e+00
-1.59173775e+00 7.23972440e-01 2.68989205e-01 -7.92398080e-02
7.31195807e-01 -2.16611847e-01 -3.07418108e-01 7.37539902e-02
-4.99830255e-03 6.36199057e-01 1.45598769e+00 -1.48877239e+00
-5.68973005e-01 -3.43483537e-01 -2.60505497e-01 4.25374746e-01
-2.95884460e-01 -4.72757816e-01 -5.93411684e-01 -9.71347868e-01
2.61669606e-01 -1.00215328e+00 -2.31548190e-01 -1.00514792e-01
-2.16294512e-01 3.10251057e-01 1.37954104e+00 -6.88401997e-01
1.00628424e+00 -2.22443438e+00 2.75012881e-01 -2.02371061e-01
3.48197341e-01 3.78109545e-01 -2.07459524e-01 4.48113352e-01
-1.63158566e-01 -4.04793084e-01 -1.51242062e-01 -2.09200725e-01
-4.48816955e-01 1.30831406e-01 -5.61734140e-01 6.79578185e-01
-1.28348097e-01 9.37463403e-01 -1.08878875e+00 -5.15889704e-01
7.07695782e-01 4.90430743e-01 -5.50233245e-01 3.15671086e-01
9.89028886e-02 7.18259573e-01 -3.27494651e-01 2.93549031e-01
1.21422648e+00 -1.81053191e-01 -1.35588557e-01 -6.17388844e-01
-2.16438636e-01 -5.51025033e-01 -1.11659968e+00 1.89938188e+00
-5.44418633e-01 4.73688990e-01 -5.45230359e-02 -8.33298862e-01
1.13518953e+00 7.85652995e-02 5.92484236e-01 -4.74871576e-01
3.60492408e-01 4.80720222e-01 -7.47916922e-02 -4.97381479e-01
2.60083884e-01 5.50931096e-02 -2.68681515e-02 3.40077519e-01
2.56828219e-02 -6.23929679e-01 -7.63599873e-02 4.37202454e-02
5.52546859e-01 6.06593862e-02 1.32048443e-01 -1.53742000e-01
8.14506531e-01 -3.66043657e-01 6.31931007e-01 2.48756856e-01
1.74684301e-01 8.90346646e-01 1.07396230e-01 -7.79737771e-01
-1.34142983e+00 -6.84875369e-01 -6.26798570e-02 3.17281991e-01
9.88432944e-01 -1.81595311e-01 -7.72463977e-01 -6.21410966e-01
-1.11074343e-01 2.33176276e-01 -5.03308356e-01 -4.92888957e-01
-8.60557914e-01 -3.63374442e-01 2.11100698e-01 2.64474571e-01
1.30409968e+00 -9.62648690e-01 -4.21511054e-01 2.39294078e-02
-2.84048706e-01 -1.25792551e+00 -1.24361396e+00 -3.74535352e-01
-1.08419871e+00 -1.10336673e+00 -9.64760542e-01 -1.26056445e+00
8.36985350e-01 1.01103210e+00 6.69933856e-01 1.99100927e-01
-1.51707143e-01 3.44914272e-02 -1.36964455e-01 2.27401048e-01
-4.09444332e-01 -1.49406135e-01 -4.10838239e-02 3.59281987e-01
-1.93517357e-01 -6.92103744e-01 -1.01104009e+00 6.22360587e-01
-1.40505934e+00 4.78984565e-01 9.60613728e-01 1.35602713e+00
2.94797152e-01 2.85170048e-01 9.09990594e-02 -7.03091919e-01
1.51846707e-01 1.25078604e-01 -8.56579959e-01 2.83879280e-01
-6.23312414e-01 1.71365533e-02 7.68676996e-01 -6.15387321e-01
-1.05259073e+00 3.05890530e-01 1.03217982e-01 -9.14567113e-01
1.81737185e-01 1.01516873e-01 -4.11977679e-01 -6.26062632e-01
3.44252884e-01 7.23179102e-01 2.86943227e-01 -2.75721759e-01
3.28732997e-01 3.11945617e-01 8.84928226e-01 -2.96228796e-01
1.41067541e+00 6.16659045e-01 4.12710793e-02 -7.48238087e-01
-5.81244230e-01 -4.41087544e-01 -7.98401892e-01 -2.26095125e-01
6.56689823e-01 -9.51110423e-01 -6.18630528e-01 1.10904610e+00
-1.25357866e+00 -1.14628345e-01 4.89125252e-02 5.37604153e-01
-7.35581040e-01 1.01473129e+00 -5.61741590e-01 -1.07065402e-01
-5.75122595e-01 -1.34010136e+00 1.41217089e+00 4.02313858e-01
5.19766986e-01 -9.69281912e-01 2.35684112e-01 4.87308651e-01
1.71619669e-01 4.23012853e-01 4.79623526e-01 2.00090483e-01
-8.61328244e-01 -4.25415859e-02 -3.83403242e-01 3.76829803e-01
4.55818266e-01 -1.42475858e-01 -7.08053410e-01 -8.06429565e-01
3.97488326e-01 -3.66533518e-01 9.64709103e-01 3.22597235e-01
1.01701844e+00 -4.30858672e-01 -4.09393847e-01 1.21826613e+00
1.61388695e+00 2.02384844e-01 8.97229254e-01 2.63678491e-01
1.15176201e+00 4.35112149e-01 6.22824192e-01 1.55193448e-01
2.61899065e-02 7.75972545e-01 4.86901402e-01 -4.89415914e-01
-3.06882888e-01 -5.19450843e-01 4.40101713e-01 7.58866191e-01
9.97297615e-02 7.35999597e-03 -3.39740872e-01 2.47706935e-01
-1.91930497e+00 -1.08031714e+00 8.72351788e-03 2.39203811e+00
7.45759368e-01 -9.89384279e-02 -1.74113020e-01 9.27844271e-02
1.02783692e+00 5.25107622e-01 -6.75791204e-01 2.12280467e-01
-1.35950267e-01 -2.48774946e-01 5.52918494e-01 5.17830849e-01
-1.09467804e+00 1.15532124e+00 5.13228178e+00 1.15079236e+00
-1.44141710e+00 -1.47336245e-01 4.75109339e-01 3.86624068e-01
-1.95600539e-01 1.19090304e-01 -6.93284035e-01 6.14438355e-01
-1.03089437e-01 -1.27217129e-01 4.80912238e-01 6.66877866e-01
1.06896581e-02 2.33486384e-01 -8.17494333e-01 1.55049062e+00
3.90592784e-01 -1.45910943e+00 2.15740472e-01 2.12125957e-01
1.23921192e+00 -3.39317292e-01 2.14447990e-01 -2.76484787e-02
1.49689451e-01 -7.03464985e-01 4.94824141e-01 1.54913679e-01
1.04087770e+00 -7.07890987e-01 6.01477861e-01 2.52816767e-01
-1.30018210e+00 -1.76279366e-01 -5.66872180e-01 3.41061324e-01
1.69063032e-01 4.54798311e-01 -3.44017386e-01 7.61785507e-01
4.82636392e-01 1.19061160e+00 -4.81071055e-01 1.08803558e+00
-1.43257603e-01 7.57332295e-02 -3.29596251e-02 5.23255467e-01
3.34933013e-01 -5.99327564e-01 4.53683943e-01 8.02579224e-01
5.61993122e-01 5.36043160e-02 4.59007651e-01 6.42092049e-01
-7.54482970e-02 -4.40866016e-02 -1.01662731e+00 3.77888978e-01
3.40716571e-01 1.40565598e+00 -7.24747002e-01 -3.21630746e-01
-3.85697335e-01 1.25712669e+00 -4.61241193e-02 3.47032428e-01
-7.90780962e-01 -3.10769320e-01 2.60384023e-01 4.66414280e-02
4.29433912e-01 -2.36252874e-01 1.14106238e-02 -1.65139055e+00
1.42764682e-02 -1.19408405e+00 2.83221304e-01 -8.13020706e-01
-1.16637540e+00 5.07256448e-01 -1.20091893e-01 -1.91596103e+00
-1.25626653e-01 -5.01271188e-01 -8.63268375e-01 5.66538215e-01
-1.44576108e+00 -1.66495585e+00 -8.39161694e-01 8.87177706e-01
7.60547817e-01 -2.02593088e-01 2.17606694e-01 1.91502839e-01
-4.54348564e-01 4.80055869e-01 2.77427852e-01 2.82599211e-01
1.07835484e+00 -8.15015912e-01 4.44322705e-01 9.67418134e-01
9.62701961e-02 5.39458871e-01 4.11030561e-01 -6.96537197e-01
-1.47833776e+00 -1.12336099e+00 1.43973947e-01 -3.09693310e-02
4.74372625e-01 -1.89882830e-01 -8.98184955e-01 5.90182304e-01
6.38316572e-01 1.55915231e-01 -3.44346054e-02 -6.54518425e-01
-4.79319096e-01 -4.93944138e-01 -8.50434244e-01 6.80533648e-01
1.07468677e+00 -2.93322802e-01 -4.71043885e-01 2.23467425e-01
8.53818595e-01 -6.76009059e-01 -6.03128910e-01 4.65467095e-01
7.09510565e-01 -1.23502827e+00 1.04430950e+00 2.78196651e-02
5.94404638e-01 -6.14565492e-01 3.73722076e-01 -1.37395942e+00
-5.35905719e-01 -1.08040440e+00 7.64525160e-02 1.14605117e+00
-3.69448215e-01 -8.56282353e-01 6.39144421e-01 1.23876743e-01
-1.59179688e-01 -4.42923754e-01 -6.15420520e-01 -8.25282156e-01
-2.23404337e-02 3.91675085e-01 6.19117558e-01 1.13370538e+00
-3.79565567e-01 5.73189318e-01 -9.03570652e-01 2.86098421e-01
1.04164672e+00 6.02467477e-01 1.27988672e+00 -9.71890450e-01
-3.65376621e-01 -4.59723145e-01 -5.26783407e-01 -1.38879371e+00
8.17158446e-02 -5.66608071e-01 -1.01960920e-01 -9.85202610e-01
2.65538365e-01 -3.08351785e-01 1.86400726e-01 1.10884815e-01
-3.87716293e-01 3.79943401e-01 4.47702371e-02 5.31764388e-01
-3.55160460e-02 8.63275349e-01 2.04679298e+00 -3.14681172e-01
-1.63029701e-01 -8.79245549e-02 -4.05140460e-01 6.16341531e-01
5.25730312e-01 -2.06239149e-01 -6.48125410e-01 -6.88628972e-01
-1.86922595e-01 2.93139696e-01 4.46764112e-01 -1.07630038e+00
3.80745351e-01 -3.11187208e-01 5.68891346e-01 -6.96604729e-01
1.94590598e-01 -9.44554090e-01 3.68358463e-01 8.86849225e-01
-1.12587951e-01 6.55191988e-02 -1.09720401e-01 6.23148024e-01
-3.79681319e-01 -1.09027728e-01 1.07790399e+00 -1.40978443e-02
-6.25124633e-01 7.45170176e-01 4.28073704e-01 -1.23993777e-01
1.03651726e+00 -3.39956939e-01 -3.85700166e-01 -3.52447301e-01
1.00162387e-01 6.22073971e-02 8.41578066e-01 6.30021930e-01
9.36386406e-01 -1.53026545e+00 -5.89040160e-01 6.71211898e-01
7.76158571e-02 2.66216725e-01 5.28466165e-01 6.89903975e-01
-8.33323359e-01 5.24945669e-02 -5.87457418e-01 -7.80574858e-01
-1.11550295e+00 1.06565320e+00 3.56475621e-01 -2.31860191e-01
-8.86960804e-01 3.96769941e-01 9.11477327e-01 -2.31109202e-01
1.14592947e-01 -1.65468916e-01 -5.30457385e-02 -3.85494828e-01
5.08232117e-01 2.47075319e-01 -2.80669421e-01 -8.49768102e-01
2.00844303e-01 1.43383765e+00 -3.10756534e-01 -4.68706004e-02
1.03798795e+00 -1.82254776e-01 -1.01489656e-01 -1.18644767e-01
1.43953586e+00 3.38208079e-02 -1.66583622e+00 -3.31959873e-01
-5.99697292e-01 -1.03674901e+00 1.21874206e-01 -1.38021827e-01
-1.68996537e+00 9.41825032e-01 6.86876893e-01 -7.51896948e-02
1.20600581e+00 -4.40293461e-01 1.27510083e+00 2.33914927e-01
4.27235395e-01 -7.00045168e-01 6.96690857e-01 1.96108788e-01
1.03455138e+00 -1.39134407e+00 1.58750072e-01 -5.49728394e-01
-6.38638496e-01 1.27025354e+00 9.04287040e-01 -6.38517201e-01
4.04579520e-01 1.43035755e-01 1.66664556e-01 -9.79822204e-02
-4.21217591e-01 1.40731871e-01 3.78453314e-01 6.64593220e-01
-2.56564785e-02 -4.13694680e-01 -2.70763755e-01 -1.14146069e-01
-8.71304572e-02 -1.40923023e-01 4.52903390e-01 4.25498456e-01
-2.70465821e-01 -1.25560665e+00 -6.04510486e-01 -8.65572169e-02
1.23963030e-02 6.59534261e-02 -2.95957327e-01 8.70568991e-01
8.93166885e-02 6.61193013e-01 -1.60280526e-01 -6.84046984e-01
1.63360000e-01 -6.57666802e-01 7.04782128e-01 -3.05180907e-01
-4.50825661e-01 4.84655112e-01 -5.90166569e-01 -4.55121815e-01
-3.31716001e-01 -3.99416119e-01 -8.89708161e-01 -2.78999031e-01
-4.96805847e-01 -1.90093666e-01 1.11827806e-01 7.12295353e-01
2.20189333e-01 2.29114499e-02 1.21626258e+00 -1.24889553e+00
-6.61868870e-01 -6.57437086e-01 -5.15782416e-01 6.60897017e-01
5.82137585e-01 -6.07095718e-01 -4.65415537e-01 3.99103165e-01] | [9.28562068939209, -2.352212905883789] |
d4b491f4-3ba5-488a-a372-e501ab401b9a | tsdae-using-transformer-based-sequential | 2104.06979 | null | https://arxiv.org/abs/2104.06979v3 | https://arxiv.org/pdf/2104.06979v3.pdf | TSDAE: Using Transformer-based Sequential Denoising Auto-Encoder for Unsupervised Sentence Embedding Learning | Learning sentence embeddings often requires a large amount of labeled data. However, for most tasks and domains, labeled data is seldom available and creating it is expensive. In this work, we present a new state-of-the-art unsupervised method based on pre-trained Transformers and Sequential Denoising Auto-Encoder (TSDAE) which outperforms previous approaches by up to 6.4 points. It can achieve up to 93.1% of the performance of in-domain supervised approaches. Further, we show that TSDAE is a strong domain adaptation and pre-training method for sentence embeddings, significantly outperforming other approaches like Masked Language Model. A crucial shortcoming of previous studies is the narrow evaluation: Most work mainly evaluates on the single task of Semantic Textual Similarity (STS), which does not require any domain knowledge. It is unclear if these proposed methods generalize to other domains and tasks. We fill this gap and evaluate TSDAE and other recent approaches on four different datasets from heterogeneous domains. | ['Iryna Gurevych', 'Nils Reimers', 'Kexin Wang'] | 2021-04-14 | null | null | null | null | ['paraphrase-identification'] | ['natural-language-processing'] | [ 1.34714425e-01 2.15744041e-02 -1.66666284e-01 -5.03599048e-01
-8.19879055e-01 -4.32094276e-01 7.48645067e-01 6.80283248e-01
-7.34881639e-01 7.23577559e-01 4.17332143e-01 -8.35420713e-02
6.91472813e-02 -6.32918000e-01 -5.17795742e-01 -3.31912756e-01
2.46827811e-01 6.60209596e-01 3.17674220e-01 -4.49708372e-01
2.24043980e-01 2.61889193e-02 -1.50164509e+00 3.65636796e-01
7.91378319e-01 8.73768449e-01 2.64261127e-01 2.86197513e-01
-4.60896879e-01 7.43260324e-01 -6.26086354e-01 -4.99690115e-01
-8.02426413e-02 -3.23862672e-01 -9.60917473e-01 -1.35847842e-02
4.96656269e-01 -1.90192699e-01 -1.29204929e-01 9.36163247e-01
7.09086120e-01 2.28572860e-01 8.09643507e-01 -9.93654490e-01
-1.21376300e+00 6.19898617e-01 -2.43461236e-01 2.13495821e-01
1.93059608e-01 -5.14017522e-01 1.20909977e+00 -1.22698343e+00
7.13553488e-01 9.85054731e-01 8.64729941e-01 6.18543565e-01
-1.10599566e+00 -3.85288060e-01 5.69986813e-02 3.79406363e-01
-1.19977403e+00 -3.02505821e-01 7.92306244e-01 -4.40059006e-01
1.33006656e+00 -2.32439652e-01 1.65706221e-02 1.52813792e+00
-1.99999571e-01 7.30558276e-01 1.28152084e+00 -7.67083645e-01
4.01349574e-01 5.43466866e-01 4.20314819e-01 3.31035852e-01
3.09531182e-01 -1.49716958e-01 -5.99644005e-01 -1.96129069e-01
1.48133576e-01 2.15167329e-02 6.92139640e-02 -3.61593455e-01
-1.05273771e+00 1.03371656e+00 7.63949007e-02 7.56688237e-01
-3.98794293e-01 -3.95975828e-01 8.10536385e-01 5.85438371e-01
9.38367784e-01 5.64288855e-01 -7.22076118e-01 -2.76199371e-01
-1.04993916e+00 1.15551271e-01 7.13786721e-01 9.05811727e-01
5.88736713e-01 -1.40392557e-02 -1.36335762e-02 1.09574819e+00
1.02666430e-01 2.43743107e-01 8.40282083e-01 -5.79961360e-01
4.70619470e-01 5.20464778e-01 5.12387492e-02 -7.71355689e-01
-2.37169296e-01 -4.86816406e-01 -6.45637453e-01 -1.22704767e-01
3.67240787e-01 -6.59670308e-02 -4.49361891e-01 1.59020126e+00
1.49069965e-01 1.40632600e-01 3.30159187e-01 7.55696595e-01
1.00844598e+00 4.84978557e-01 2.34563768e-01 -2.01256834e-02
1.39723289e+00 -1.23227501e+00 -8.78536224e-01 -4.86707240e-01
9.10461724e-01 -1.01240969e+00 1.31664157e+00 3.74859452e-01
-8.63284349e-01 -7.18206465e-01 -1.09192264e+00 -3.22668076e-01
-7.10333169e-01 3.85474175e-01 3.80415589e-01 6.01249099e-01
-9.06676173e-01 6.60961866e-01 -5.31414568e-01 -9.16456938e-01
2.44135201e-01 -8.17033276e-03 -5.35519600e-01 -2.67953068e-01
-1.48801196e+00 1.30975306e+00 4.31528568e-01 -6.15536034e-01
-6.51352584e-01 -8.18491638e-01 -9.60013747e-01 1.26802502e-02
1.70439512e-01 -4.97021586e-01 1.34778833e+00 -8.76288414e-01
-1.53253388e+00 1.11908126e+00 -3.66183192e-01 -7.80897319e-01
1.21242218e-01 -5.87727308e-01 -6.79948330e-01 6.06150776e-02
2.24771813e-01 3.91820431e-01 6.55226827e-01 -1.01461506e+00
-3.28482717e-01 -3.47049475e-01 3.30603536e-04 1.28204256e-01
-1.18315780e+00 2.59293050e-01 6.02414012e-02 -8.49152088e-01
-4.29356575e-01 -6.00783646e-01 -1.20825335e-01 -1.06394440e-01
7.59296715e-02 -5.19091427e-01 8.90560567e-01 -7.10858762e-01
1.41012406e+00 -2.26497722e+00 -5.69440611e-03 -4.70716536e-01
1.97989736e-02 5.79321921e-01 -3.19244295e-01 9.30972278e-01
-2.01183155e-01 -3.25642228e-02 -4.08380359e-01 -7.09149420e-01
3.54274988e-01 2.30025262e-01 -4.25374538e-01 1.87397972e-01
4.51775938e-01 6.69378161e-01 -1.02268410e+00 -5.03380299e-01
2.28924334e-01 4.54072028e-01 -4.11203563e-01 2.90215522e-01
-5.19536808e-02 1.90679446e-01 -2.50087410e-01 3.64858568e-01
7.28794098e-01 -2.38847703e-01 2.04005405e-01 -1.00046642e-01
-8.15092847e-02 6.17463708e-01 -9.35460329e-01 2.10082936e+00
-7.88565338e-01 7.20519543e-01 -3.28032374e-01 -1.63580596e+00
1.15381324e+00 4.14639831e-01 4.04219836e-01 -7.66611934e-01
1.88744470e-01 3.91152024e-01 -3.06040794e-01 -6.23636186e-01
6.38666987e-01 -3.25816602e-01 -2.20018297e-01 5.30808151e-01
6.23080373e-01 1.07802257e-01 2.62477666e-01 2.03277960e-01
1.22980642e+00 2.67132232e-03 4.75506157e-01 -3.55034292e-01
4.73743141e-01 1.50711117e-02 2.50139862e-01 4.73723620e-01
-2.51856238e-01 6.73596919e-01 3.38673413e-01 -2.11288810e-01
-1.30712068e+00 -8.75220656e-01 -3.18670243e-01 1.27910650e+00
-9.79178622e-02 -5.75887859e-01 -7.93865144e-01 -8.65908206e-01
-1.17330402e-01 1.03736973e+00 -6.32873297e-01 -2.05648258e-01
-1.52190208e-01 -6.75688326e-01 5.02281010e-01 7.42230475e-01
3.36479098e-01 -8.30411971e-01 -1.13829665e-01 3.78209352e-01
-1.99901491e-01 -1.49670458e+00 -2.63704866e-01 1.44040212e-01
-1.06548476e+00 -7.24500835e-01 -7.84852266e-01 -1.01373112e+00
4.27337974e-01 2.21784011e-01 1.40643966e+00 -4.05070603e-01
2.30914019e-02 2.06682384e-01 -8.85168135e-01 -5.71151495e-01
-3.42660695e-01 3.26421559e-01 2.39089951e-01 -4.95442711e-02
1.16430330e+00 -7.28550375e-01 -2.28615940e-01 1.35256410e-01
-9.07504201e-01 -5.45696139e-01 5.82246125e-01 1.04031432e+00
4.49083686e-01 -1.73746608e-02 9.43126023e-01 -1.10361767e+00
9.36407387e-01 -7.14126229e-01 -1.56290650e-01 2.74743468e-01
-7.08681345e-01 2.74882942e-01 8.56548429e-01 -5.03933132e-01
-1.04418981e+00 -6.92852437e-02 -3.11002821e-01 -2.25879267e-01
-4.06396240e-01 5.61296403e-01 7.33839795e-02 3.32061887e-01
9.00844514e-01 1.52910635e-01 -1.27353445e-01 -8.20609093e-01
3.54451537e-01 1.07334769e+00 1.57104716e-01 -4.93837655e-01
6.65608943e-01 4.17282104e-01 -5.08134365e-01 -9.70983565e-01
-1.09765029e+00 -6.77040875e-01 -7.38105714e-01 2.30170488e-01
7.97230124e-01 -9.69277322e-01 1.59518234e-02 3.59280258e-02
-1.28260851e+00 -6.76823407e-02 -3.99949342e-01 7.11702704e-01
-2.59662032e-01 7.15311050e-01 -3.89705747e-01 -4.81532693e-01
-3.95052552e-01 -6.80918396e-01 9.84043121e-01 -2.34100983e-01
-4.86954719e-01 -1.35700345e+00 3.12553406e-01 3.84368747e-01
6.70293868e-01 -4.65491861e-02 7.06342578e-01 -1.32015979e+00
2.83354759e-01 -3.43681991e-01 -1.05887420e-01 8.59334290e-01
6.10306114e-02 -3.89204025e-01 -1.14067376e+00 -2.92736292e-01
2.84991920e-01 -7.65720785e-01 8.01914036e-01 1.23016179e-01
9.86679018e-01 9.00774524e-02 -3.16773467e-02 1.01681426e-01
1.42152667e+00 -1.57103181e-01 4.90283579e-01 5.89384973e-01
2.55474567e-01 6.89840615e-01 7.94896305e-01 4.64763254e-01
3.83470327e-01 6.46397531e-01 9.91199017e-02 -1.64809953e-02
-1.93945929e-01 -2.82004595e-01 5.42956889e-01 1.36740279e+00
1.04832627e-01 -1.80810273e-01 -7.74510920e-01 8.73780787e-01
-1.81392968e+00 -8.63601625e-01 -1.29313990e-01 2.07085443e+00
9.01496649e-01 6.68842047e-02 -3.95072289e-02 2.45380268e-01
7.43688345e-01 1.36302516e-01 -2.08373904e-01 -6.71350837e-01
-1.24518633e-01 7.31406569e-01 3.14880937e-01 3.12404990e-01
-1.28756845e+00 8.67257476e-01 6.38176107e+00 1.00104594e+00
-7.75120735e-01 7.04255641e-01 1.06881186e-01 1.49624169e-01
-1.83174849e-01 -2.82834657e-02 -7.38845766e-01 3.59098762e-01
1.16852510e+00 -1.55176282e-01 -9.62494127e-03 1.18183112e+00
-8.66398960e-02 2.03556955e-01 -1.14382660e+00 8.08729649e-01
4.59555566e-01 -9.47159827e-01 -2.25529611e-01 -2.64238417e-01
8.28659058e-01 1.09491661e-01 2.37057507e-02 5.87245703e-01
2.49740943e-01 -9.07968998e-01 2.58794010e-01 1.42283440e-01
6.67782366e-01 -6.06053531e-01 1.17474151e+00 3.19053471e-01
-8.54734182e-01 3.87834571e-02 -7.13070452e-01 -2.63132870e-01
2.10643977e-01 8.56849909e-01 -6.70693159e-01 6.86103642e-01
6.66411042e-01 1.00245750e+00 -5.51732659e-01 6.81353152e-01
-2.94799894e-01 8.37795734e-01 -2.06166163e-01 -2.18592197e-01
3.38150710e-01 1.27446115e-01 1.81153983e-01 1.45857894e+00
4.61250454e-01 -2.43780404e-01 2.23620869e-02 5.99764943e-01
-1.33001357e-01 3.57202828e-01 -8.09707463e-01 -3.41645837e-01
6.06306255e-01 1.05812025e+00 -2.55979270e-01 -5.10732532e-01
-8.50613713e-01 1.18029261e+00 5.25516629e-01 1.61076218e-01
-8.15271080e-01 -7.38393605e-01 5.26505411e-01 1.77274439e-02
6.31217360e-01 -2.18258679e-01 -2.46319637e-01 -1.39333975e+00
2.76901245e-01 -7.92696238e-01 3.55818868e-01 -4.22990412e-01
-2.00772858e+00 7.91135252e-01 -1.91964224e-01 -1.54578030e+00
-3.18598896e-01 -7.83754885e-01 -3.18155408e-01 6.47124350e-01
-1.85692179e+00 -1.02877343e+00 -1.22961625e-01 5.42097569e-01
7.04168081e-01 -5.57763636e-01 1.35740209e+00 7.14054942e-01
-2.87049353e-01 7.64999151e-01 6.15318239e-01 1.46453947e-01
1.24056220e+00 -1.21461225e+00 3.70450020e-01 6.45275295e-01
3.40780824e-01 5.97115040e-01 7.06725478e-01 -2.58403063e-01
-9.65090752e-01 -1.02616763e+00 1.69994545e+00 -6.53694868e-01
1.01614606e+00 -5.14471650e-01 -1.03181088e+00 5.42218804e-01
7.46144474e-01 5.93720041e-02 1.03623176e+00 2.95022398e-01
-6.01407588e-01 -1.40162498e-01 -1.01675689e+00 1.92008436e-01
9.02138472e-01 -7.18019724e-01 -1.24015999e+00 5.13491869e-01
7.66199589e-01 2.49026623e-02 -9.95280325e-01 2.15409100e-01
1.07553899e-01 -9.56720173e-01 1.00968528e+00 -7.81389177e-01
7.89668322e-01 -3.04433554e-02 -3.54097217e-01 -1.46810794e+00
-2.19207704e-01 -7.23812804e-02 -1.04576603e-01 1.53016138e+00
3.51213306e-01 -6.50864720e-01 5.55782259e-01 1.33883074e-01
-2.03327924e-01 -5.38168490e-01 -7.87129402e-01 -1.13469946e+00
4.13879961e-01 -4.66160268e-01 3.92536789e-01 1.42199945e+00
1.02703065e-01 5.95292985e-01 -2.82570571e-01 2.93696299e-02
5.68149149e-01 -7.86694214e-02 3.92645627e-01 -1.18534803e+00
-2.00394034e-01 -1.39938429e-01 -4.82301205e-01 -8.94350290e-01
5.65705240e-01 -1.04133129e+00 -2.78801888e-01 -1.71616530e+00
2.69573238e-02 -8.72583985e-02 -4.84017909e-01 2.72108197e-01
-8.12127590e-02 1.64606974e-01 -1.58684611e-01 -6.92733303e-02
-7.22023129e-01 8.21939647e-01 7.04607010e-01 -1.83096245e-01
2.00356767e-01 -3.65272552e-01 -6.37598693e-01 7.31406987e-01
9.74440515e-01 -7.73660183e-01 -5.19391477e-01 -7.18595445e-01
-8.25809911e-02 -5.33727288e-01 1.63955107e-01 -9.79879797e-01
2.47732431e-01 1.77243248e-01 4.91151307e-03 -3.24625850e-01
2.62967527e-01 -8.25032830e-01 -4.92119074e-01 1.78810462e-01
-4.73379284e-01 1.33259356e-01 1.66269287e-01 4.86733437e-01
-8.52234542e-01 -7.58735538e-01 5.35844266e-01 -8.56415555e-02
-9.32429910e-01 -1.20073915e-01 -4.00863469e-01 5.14157116e-01
9.01905179e-01 4.00442518e-02 -4.68862057e-02 -1.95447549e-01
-6.47858858e-01 -1.35509148e-01 3.03623497e-01 6.61799967e-01
5.16233325e-01 -1.60246587e+00 -8.42229307e-01 1.84588488e-02
5.79226613e-01 -3.31484288e-01 3.55380774e-01 5.68787813e-01
-2.06877634e-01 4.58288074e-01 -4.68199775e-02 -3.14178377e-01
-1.16543007e+00 7.38110363e-01 -1.31087929e-01 -4.85367119e-01
-4.03495669e-01 7.56051004e-01 -2.46262074e-01 -7.67639756e-01
2.07685173e-01 -2.91549098e-02 -3.01521271e-01 2.94527411e-01
5.86042821e-01 2.35679999e-01 3.36207151e-01 -4.68266726e-01
-4.17026341e-01 6.07171178e-01 -1.54232427e-01 -5.63183539e-02
1.57216084e+00 2.18421075e-04 -1.74892277e-01 5.06135345e-01
1.51049709e+00 -1.88348040e-01 -6.07282877e-01 -7.00560749e-01
2.54888654e-01 -3.85927111e-01 5.92933334e-02 -5.65609217e-01
-6.59148872e-01 1.41541195e+00 5.80768049e-01 1.60579070e-01
1.12132514e+00 3.97917964e-02 9.96651649e-01 5.78034520e-01
2.89671987e-01 -1.55911827e+00 3.77530485e-01 7.50342011e-01
6.54023588e-01 -1.46382713e+00 -1.21639103e-01 -1.60754144e-01
-9.32821631e-01 1.16689050e+00 5.09216845e-01 -2.79907823e-01
7.26144791e-01 2.30736881e-01 6.11927584e-02 -3.89295593e-02
-7.14023590e-01 -3.29152018e-01 2.14025542e-01 7.35513806e-01
8.86895299e-01 -2.11599290e-01 -7.36680388e-01 9.46842253e-01
-9.50647742e-02 1.86548039e-01 2.13891312e-01 9.58003283e-01
-4.24889416e-01 -1.66766417e+00 -6.43341616e-02 1.28532708e-01
-5.65028846e-01 -3.68122488e-01 -4.18608397e-01 7.04721093e-01
1.43064722e-01 9.85732675e-01 1.86891705e-02 -3.60885113e-01
4.53592449e-01 4.75367367e-01 3.22239459e-01 -8.56046259e-01
-4.37960625e-01 -3.63564402e-01 3.33004206e-01 -2.15781868e-01
-7.77551055e-01 -6.51117325e-01 -9.34536636e-01 -2.16906339e-01
-2.07439065e-01 1.88277364e-01 5.44332445e-01 8.81974697e-01
5.37793219e-01 4.26203221e-01 4.90748763e-01 -4.41584796e-01
-8.60845923e-01 -1.27991486e+00 -5.57869732e-01 8.04844618e-01
-4.39209491e-02 -7.11366415e-01 -3.88711452e-01 1.21666171e-01] | [10.582518577575684, 8.772275924682617] |
e4309687-5cbb-4a7c-822a-c18c478e58f0 | lsmi-sinkhorn-semi-supervised-squared-loss | 1909.02373 | null | https://arxiv.org/abs/1909.02373v3 | https://arxiv.org/pdf/1909.02373v3.pdf | LSMI-Sinkhorn: Semi-supervised Mutual Information Estimation with Optimal Transport | Estimating mutual information is an important statistics and machine learning problem. To estimate the mutual information from data, a common practice is preparing a set of paired samples $\{(\mathbf{x}_i,\mathbf{y}_i)\}_{i=1}^n \stackrel{\mathrm{i.i.d.}}{\sim} p(\mathbf{x},\mathbf{y})$. However, in many situations, it is difficult to obtain a large number of data pairs. To address this problem, we propose the semi-supervised Squared-loss Mutual Information (SMI) estimation method using a small number of paired samples and the available unpaired ones. We first represent SMI through the density ratio function, where the expectation is approximated by the samples from marginals and its assignment parameters. The objective is formulated using the optimal transport problem and quadratic programming. Then, we introduce the Least-Squares Mutual Information with Sinkhorn (LSMI-Sinkhorn) algorithm for efficient optimization. Through experiments, we first demonstrate that the proposed method can estimate the SMI without a large number of paired samples. Then, we show the effectiveness of the proposed LSMI-Sinkhorn algorithm on various types of machine learning problems such as image matching and photo album summarization. Code can be found at https://github.com/csyanbin/LSMI-Sinkhorn. | ['Yao-Hung Hubert Tsai', 'Yanbin Liu', 'Makoto Yamada', 'Yi Yang', 'Ruslan Salakhutdinov', 'Tam Le'] | 2019-09-05 | null | null | null | null | ['mutual-information-estimation'] | ['methodology'] | [ 3.60443801e-01 -1.75950006e-01 -4.59227450e-02 -6.06471956e-01
-1.37852585e+00 -4.07430977e-01 1.15336597e-01 4.15518740e-03
-4.54673380e-01 1.06791937e+00 -3.17431688e-01 -1.85634524e-01
-4.82780993e-01 -6.59835637e-01 -8.63623559e-01 -1.01365936e+00
-2.48684082e-02 4.80553865e-01 -1.45052612e-01 3.56160969e-01
2.83012301e-01 9.76005942e-02 -1.55267370e+00 -1.03377379e-01
1.23760986e+00 1.16628993e+00 3.86118263e-01 5.61715841e-01
1.94110647e-02 6.29502177e-01 -5.28278410e-01 -4.95736957e-01
4.36448961e-01 -7.64609098e-01 -7.92671084e-01 -3.39888155e-01
3.79602671e-01 -2.30807230e-01 -1.47294059e-01 1.35333490e+00
6.04412019e-01 2.86330491e-01 9.09130633e-01 -1.43460011e+00
-2.77609676e-01 5.00568688e-01 -1.22183263e+00 -6.17981656e-03
8.89948979e-02 -1.13793753e-01 9.95577276e-01 -7.63589025e-01
2.95056462e-01 8.85110736e-01 4.65408593e-01 1.66274875e-01
-1.13556874e+00 -1.16795421e+00 -3.23290378e-01 -2.06877831e-02
-1.80746830e+00 -2.36829177e-01 5.33245802e-01 -2.19545841e-01
2.68007487e-01 7.78401792e-01 6.38967529e-02 2.10177854e-01
-1.50667548e-01 1.15172100e+00 1.32261312e+00 -4.11386698e-01
-9.28333923e-02 3.29669833e-01 1.36466190e-01 6.35384262e-01
6.04474284e-02 2.59722080e-02 -4.80464190e-01 -2.84828871e-01
5.29364228e-01 1.01948105e-01 -1.48693502e-01 9.03808698e-02
-1.04703856e+00 6.04758561e-01 3.54417026e-01 1.58331081e-01
-2.15422679e-02 8.18729177e-02 -2.86725387e-02 3.02404523e-01
5.34472525e-01 7.98172578e-02 -4.64030117e-01 2.64952183e-02
-1.20736253e+00 1.39562383e-01 7.12010622e-01 1.08647609e+00
1.20561445e+00 -4.39627469e-01 4.57747094e-02 1.03557050e+00
4.08558726e-01 8.83146644e-01 1.95190366e-02 -1.08690083e+00
5.68289697e-01 3.31363350e-01 8.71657282e-02 -1.06230772e+00
-9.69453380e-02 -2.88828999e-01 -1.12330711e+00 -5.28453626e-02
5.43591797e-01 -2.97959715e-01 -4.61617619e-01 1.86337495e+00
2.79521853e-01 1.99018151e-01 -2.05054671e-01 7.69962251e-01
7.57440746e-01 7.44982064e-01 -2.03339368e-01 -5.66855073e-01
9.55788672e-01 -5.91190875e-01 -3.54094177e-01 -1.58160135e-01
4.93026584e-01 -9.90679979e-01 7.10351527e-01 3.03189039e-01
-1.22633743e+00 -2.13832557e-01 -7.04415381e-01 5.10237999e-02
3.07662617e-02 4.34245169e-01 5.06201148e-01 3.61286610e-01
-8.06891680e-01 5.61609924e-01 -4.51608956e-01 -3.21642235e-02
4.13560897e-01 6.52523875e-01 -5.93828380e-01 -2.03509465e-01
-9.10632551e-01 3.05629879e-01 2.13245600e-01 1.11683249e-01
-4.71519202e-01 -6.98354185e-01 -7.39715338e-01 -2.30685294e-01
1.72057688e-01 -4.55348641e-01 7.36275792e-01 -9.20859516e-01
-1.06116688e+00 1.14079714e+00 -5.22619009e-01 -5.25171831e-02
3.96114379e-01 8.81714001e-03 6.57404810e-02 -7.68587887e-02
3.15984696e-01 5.84518492e-01 5.18607199e-01 -1.31229615e+00
-5.86888433e-01 -6.97014987e-01 -4.67998415e-01 1.63784489e-01
-1.62493452e-01 2.95674384e-01 -5.37119806e-01 -6.07035697e-01
4.39699948e-01 -8.83763611e-01 -6.73181415e-02 -2.09113061e-02
-7.90077806e-01 8.86024684e-02 3.38455051e-01 -9.68987048e-01
1.22569096e+00 -2.28137589e+00 -4.91313338e-02 5.23394585e-01
1.49892792e-01 1.15781091e-01 1.35906572e-02 3.59642446e-01
-1.08357094e-01 1.01114422e-01 -6.83225095e-01 -3.15703750e-01
-8.27450007e-02 -1.17425695e-01 2.12514907e-01 5.78705966e-01
-2.76390851e-01 4.60936993e-01 -8.23408425e-01 -6.19864225e-01
1.69543549e-01 2.64946640e-01 -3.29076976e-01 3.67555171e-01
-5.68767078e-03 7.14866638e-01 -2.60003567e-01 5.53175390e-01
1.37315989e+00 -4.04138297e-01 1.85299590e-01 -4.31673169e-01
-6.46608397e-02 -1.84116378e-01 -1.49551499e+00 1.34449637e+00
-2.91468203e-01 4.58133131e-01 1.08757287e-01 -1.19347429e+00
1.00363123e+00 -7.93473050e-02 7.77616560e-01 -6.05563462e-01
3.43231648e-01 3.86718482e-01 -1.77810624e-01 -2.81911314e-01
1.29823402e-01 -3.12539577e-01 -4.88821380e-02 5.62576473e-01
-1.27588630e-01 -1.17939644e-01 1.42393470e-01 1.51609972e-01
9.90355849e-01 -2.03239515e-01 2.69513607e-01 -1.01748012e-01
5.26251674e-01 -3.93696636e-01 7.14660466e-01 6.63584888e-01
-1.31638408e-01 6.88034534e-01 5.05207896e-01 2.54670620e-01
-9.04917598e-01 -1.22654116e+00 -3.26523125e-01 7.53277838e-01
3.21737260e-01 -8.56062106e-04 -6.94426060e-01 -4.71550912e-01
-1.28812073e-02 7.18147635e-01 -3.22590381e-01 2.42956746e-02
-7.89553598e-02 -1.08444142e+00 3.98615539e-01 1.63674235e-01
6.84567273e-01 -6.03321373e-01 1.04070187e-01 -3.51032645e-01
-5.09342492e-01 -6.15971446e-01 -6.45586371e-01 3.85681204e-02
-6.41657948e-01 -9.47442651e-01 -6.56113505e-01 -7.11384237e-01
1.00667024e+00 3.11051935e-01 8.90735507e-01 4.42616120e-02
-4.29564565e-01 4.13554301e-03 -1.91472068e-01 -2.97815770e-01
5.72932186e-03 -3.23951125e-01 -6.74469918e-02 1.07148357e-01
3.69794309e-01 -6.81226790e-01 -7.82631934e-01 6.79268479e-01
-8.78301084e-01 1.19906515e-02 5.71286321e-01 9.14183378e-01
1.05757320e+00 1.94361329e-01 3.14380646e-01 -7.81049967e-01
2.15726733e-01 -4.99340922e-01 -7.44238675e-01 3.87141228e-01
-3.51378590e-01 -1.07170865e-02 2.48073280e-01 -4.74837422e-02
-1.01168513e+00 -4.75338986e-03 -5.87273799e-02 -3.63306165e-01
9.99113098e-02 4.28199679e-01 -4.20668095e-01 -5.15808202e-02
2.50655621e-01 2.29065090e-01 2.88677990e-01 -4.73086745e-01
2.78401434e-01 1.09073043e+00 4.41489428e-01 -5.25735617e-01
5.59775651e-01 4.84000474e-01 9.67252553e-02 -7.71698534e-01
-9.63408709e-01 -5.73105216e-01 -2.78040320e-01 -1.09242737e-01
7.08893955e-01 -8.15237761e-01 -1.11197639e+00 6.85000122e-01
-8.05537522e-01 -3.14764827e-01 -1.02312461e-01 7.72700489e-01
-5.31713009e-01 5.51793218e-01 -3.82286966e-01 -1.07824492e+00
-4.04525548e-01 -1.06009388e+00 9.68490541e-01 4.13034737e-01
-6.23706030e-03 -8.59850705e-01 1.43951057e-02 7.54634738e-01
-1.48925677e-01 1.79023027e-01 5.96924961e-01 -6.15172625e-01
-6.04986429e-01 -3.69342297e-01 -6.44420981e-01 6.98861480e-01
2.14340284e-01 -2.27041058e-02 -7.25568354e-01 -2.30138481e-01
-9.84511524e-02 -2.40089238e-01 7.76633739e-01 6.45896554e-01
1.47214389e+00 -4.63142782e-01 -3.47821981e-01 7.27743983e-01
1.45260632e+00 4.35335236e-03 7.62972116e-01 -1.80594593e-01
5.17278016e-01 5.46833336e-01 8.84025693e-01 8.38831425e-01
4.77680802e-01 3.81217122e-01 2.07076997e-01 -1.99906398e-02
2.40997523e-01 -1.89037621e-01 1.30391315e-01 8.94371867e-01
8.08272362e-02 -3.24987143e-01 -7.23497450e-01 5.17052531e-01
-1.76496613e+00 -9.89739835e-01 -4.03586477e-01 2.79433227e+00
1.04715729e+00 -4.43551898e-01 -9.75836739e-02 1.29092941e-02
8.96206975e-01 -1.92916915e-01 -4.87302363e-01 1.56377614e-01
-2.54755735e-01 2.88560927e-01 7.47726381e-01 6.33647859e-01
-9.18413103e-01 5.46402216e-01 4.91820240e+00 1.26458859e+00
-7.31003284e-01 1.58128999e-02 9.95829105e-01 -9.26392078e-02
-4.42723334e-01 3.29966992e-01 -6.85741127e-01 9.53462541e-01
5.36613941e-01 4.95064296e-02 6.13591909e-01 2.94732243e-01
1.17526308e-01 -7.46680796e-01 -8.51068318e-01 1.18615210e+00
1.36869177e-01 -8.44440937e-01 -3.72924805e-01 3.94627973e-02
9.34309721e-01 -1.75125763e-01 2.26119533e-01 -1.00814670e-01
4.45884615e-01 -9.85202312e-01 2.76966035e-01 8.32037866e-01
9.57007110e-01 -7.18816280e-01 7.82184005e-01 5.86454690e-01
-9.60720301e-01 1.91195294e-01 -2.74555385e-01 7.87579939e-02
6.01840653e-02 1.27395439e+00 -4.75983322e-01 7.90361166e-01
8.51867855e-01 3.77155542e-01 -1.51744097e-01 1.14120150e+00
-2.51555108e-02 5.20790517e-01 -7.67352819e-01 3.91187333e-02
-3.04079145e-01 -7.24141359e-01 4.91798311e-01 6.94675267e-01
6.14604533e-01 3.34003121e-01 7.48468414e-02 7.58707881e-01
-2.49070331e-01 3.03532749e-01 -3.22662085e-01 5.02274372e-03
7.11803555e-01 1.06474781e+00 -5.31171143e-01 9.47368704e-03
-2.40707904e-01 1.00970900e+00 3.59200448e-01 2.29019716e-01
-8.29315364e-01 -6.28519356e-01 5.43663442e-01 -4.81068343e-02
5.37911803e-02 4.05092277e-02 -3.96468848e-01 -8.86343837e-01
1.71347082e-01 -7.17473626e-01 5.64049304e-01 -8.26638341e-01
-1.50231874e+00 1.05310842e-01 2.59933919e-02 -1.28481805e+00
1.23346476e-02 -3.16368073e-01 -5.00737250e-01 1.11337841e+00
-1.17850173e+00 -1.00805783e+00 -2.38277167e-01 4.05168951e-01
-1.51916906e-01 -9.72261503e-02 5.38866937e-01 6.51423514e-01
-6.53591514e-01 9.80975509e-01 6.51594222e-01 1.05940163e-01
8.34078670e-01 -9.84003723e-01 -2.41795182e-01 5.10781586e-01
-1.64696932e-01 5.93597889e-01 6.66939378e-01 -6.57988191e-01
-1.13598728e+00 -7.69673884e-01 9.86308038e-01 -1.21529117e-01
4.55700696e-01 -3.13692307e-03 -6.57073438e-01 6.35359883e-01
-1.03626810e-02 7.54242465e-02 1.00818384e+00 -2.14208201e-01
-3.15238237e-02 -3.35136831e-01 -1.56973207e+00 4.46116745e-01
8.36477995e-01 -3.66440743e-01 1.33177221e-01 5.05514264e-01
3.05434018e-01 -2.80745268e-01 -1.11341846e+00 5.85813165e-01
5.38381159e-01 -1.10060215e+00 9.25772130e-01 -1.74742609e-01
4.07255411e-01 -4.07840818e-01 -4.86877441e-01 -9.42502141e-01
2.58286804e-01 -6.36418521e-01 4.24381614e-01 1.53599620e+00
5.86757898e-01 -6.61991537e-01 7.74839759e-01 8.48900616e-01
4.03080314e-01 -6.83599353e-01 -1.02118230e+00 -5.96434295e-01
7.88537879e-03 -4.20362294e-01 4.27113593e-01 8.02719593e-01
-7.80837089e-02 -7.85380527e-02 -6.98242724e-01 1.40199751e-01
9.87056673e-01 1.59456551e-01 8.59629273e-01 -9.34375644e-01
-5.25007784e-01 -1.88520818e-03 -3.10270816e-01 -9.83134627e-01
9.10164714e-02 -1.05825245e+00 2.38638997e-01 -1.39828968e+00
8.48280191e-01 -9.13443208e-01 -3.44006121e-01 3.46224964e-01
-4.21096891e-01 2.56015509e-01 8.97099227e-02 2.22785622e-01
-5.69596469e-01 5.63668609e-01 1.13267052e+00 -4.24028076e-02
-3.08268536e-02 2.80293852e-01 -6.70497298e-01 5.47248840e-01
8.21334302e-01 -5.85558116e-01 -2.35974357e-01 -2.75632888e-01
2.55293190e-01 3.64222467e-01 3.43120396e-01 -7.90642917e-01
2.83116490e-01 -2.19241142e-01 2.43647531e-01 -7.90655911e-01
2.99977213e-01 -5.99265158e-01 4.32530642e-01 9.89532024e-02
-1.67075008e-01 -3.05255592e-01 -8.97805169e-02 3.51571232e-01
-2.98032194e-01 -4.56516415e-01 8.23677421e-01 -6.65199086e-02
-1.26611918e-01 4.65696007e-01 1.72210291e-01 1.58946574e-01
1.03315485e+00 4.31952067e-02 -1.83333457e-01 -5.12398124e-01
-4.37441796e-01 4.61409867e-01 3.85235250e-01 -2.66734362e-01
5.57041407e-01 -1.42103171e+00 -8.33675027e-01 1.04266316e-01
4.39275205e-02 1.65190682e-01 5.97281814e-01 1.31431305e+00
-2.83150733e-01 -8.60126689e-02 6.86334446e-02 -5.34521341e-01
-1.36134553e+00 1.44897923e-01 7.46561289e-02 -2.45242357e-01
1.84086189e-01 1.22613132e+00 2.33815387e-01 -8.55466425e-01
3.69282290e-02 2.59553820e-01 3.21678340e-01 5.53343780e-02
4.07154590e-01 7.44441032e-01 -2.07591236e-01 -6.81593657e-01
-2.83873796e-01 6.37530208e-01 2.69326046e-02 -2.18772277e-01
1.18697345e+00 -3.92914593e-01 -7.97606707e-01 4.06683058e-01
1.83457744e+00 6.37938501e-03 -1.03210163e+00 -3.10921967e-01
-3.86329561e-01 -8.61273587e-01 -2.08023667e-01 -6.25366628e-01
-1.13423049e+00 7.35939860e-01 6.94663823e-01 -2.21856400e-01
1.18039274e+00 6.26374856e-02 7.83859193e-01 7.02419505e-02
3.84167016e-01 -1.07286942e+00 -7.69309327e-02 1.91695526e-01
5.79343557e-01 -1.54295230e+00 1.60624638e-01 -5.20042241e-01
-5.61143935e-01 5.70474803e-01 3.95869702e-01 1.26899004e-01
9.97557282e-01 2.48929650e-01 -1.06696701e-02 -8.96566287e-02
-1.19210921e-01 -2.40967259e-01 2.30722010e-01 1.99795276e-01
6.03827477e-01 2.68042147e-01 -5.70297241e-01 5.69875538e-01
-8.66964683e-02 -5.09206876e-02 -7.27932155e-02 7.62935519e-01
-3.03539336e-01 -1.26859140e+00 -4.13258761e-01 9.58799720e-01
-4.35311764e-01 -1.59906864e-01 -2.13243201e-01 3.82111549e-01
2.15440571e-01 9.87065017e-01 1.01229191e-01 -3.89251024e-01
-1.06615365e-01 -1.58937141e-01 5.83913624e-01 -1.99338228e-01
-2.06730023e-01 -4.81485836e-02 -2.13409469e-01 -2.11412802e-01
-4.27364409e-01 -1.03538036e+00 -1.39506471e+00 -5.23650706e-01
-6.98797047e-01 1.56135544e-01 5.84710896e-01 8.60315084e-01
1.08923368e-01 1.13415010e-02 1.03603470e+00 -4.90682721e-01
-5.01490533e-01 -8.77610922e-01 -9.66006041e-01 3.72937649e-01
-1.15147221e-03 -4.53032941e-01 -5.48439085e-01 -1.29139438e-01] | [9.245649337768555, 3.800542116165161] |
18b66f67-c6c9-4c32-8618-39c0588d4552 | measuring-bias-in-ai-models-with-application | 2304.13680 | null | https://arxiv.org/abs/2304.13680v2 | https://arxiv.org/pdf/2304.13680v2.pdf | Measuring Bias in AI Models: An Statistical Approach Introducing N-Sigma | The new regulatory framework proposal on Artificial Intelligence (AI) published by the European Commission establishes a new risk-based legal approach. The proposal highlights the need to develop adequate risk assessments for the different uses of AI. This risk assessment should address, among others, the detection and mitigation of bias in AI. In this work we analyze statistical approaches to measure biases in automatic decision-making systems. We focus our experiments in face recognition technologies. We propose a novel way to measure the biases in machine learning models using a statistical approach based on the N-Sigma method. N-Sigma is a popular statistical approach used to validate hypotheses in general science such as physics and social areas and its application to machine learning is yet unexplored. In this work we study how to apply this methodology to develop new risk assessment frameworks based on bias analysis and we discuss the main advantages and drawbacks with respect to other popular statistical tests. | ['Javier Ortega-Garcia', 'Julian Fierrez', 'Aythami Morales', 'Ignacio Serna', 'Daniel DeAlcala'] | 2023-04-26 | null | null | null | null | ['face-recognition'] | ['computer-vision'] | [ 4.46758151e-01 5.07399619e-01 -3.31232995e-01 -6.48430645e-01
-2.81770945e-01 -1.97890937e-01 9.81899202e-01 2.98450738e-01
-6.03643119e-01 8.27308714e-01 1.95308641e-01 -4.41428751e-01
-6.36538744e-01 -9.66520727e-01 -4.31164414e-01 -5.98111987e-01
2.01260641e-01 5.11204958e-01 -8.28914195e-02 -6.20276779e-02
6.60703123e-01 8.82619262e-01 -1.84673977e+00 2.20622823e-01
9.81587708e-01 8.06231499e-01 -6.62821591e-01 1.79160044e-01
-8.29507485e-02 4.95153010e-01 -6.59209251e-01 -7.49594510e-01
1.30983621e-01 -5.89812160e-01 -7.13293672e-01 -6.44528329e-01
3.23721856e-01 -6.54940605e-02 6.15272164e-01 1.27831542e+00
4.31642115e-01 -2.07238674e-01 1.35763848e+00 -1.32579136e+00
-4.05809522e-01 8.29395771e-01 -3.58453333e-01 1.79371655e-01
3.68298203e-01 -9.04325023e-02 5.33678651e-01 -3.95730793e-01
2.77957529e-01 1.57526779e+00 3.56408477e-01 7.55487084e-01
-1.07007253e+00 -1.09849107e+00 -1.15541304e-02 3.24498057e-01
-9.32816207e-01 -1.66918352e-01 6.86420560e-01 -8.19080770e-01
4.21089888e-01 4.25238371e-01 5.38768470e-01 1.11356807e+00
5.27928472e-01 2.63527840e-01 1.79464304e+00 -8.27881336e-01
6.57319188e-01 5.17099142e-01 4.82114673e-01 4.18346584e-01
1.03855991e+00 5.15700340e-01 -4.22691673e-01 -4.34580356e-01
2.88410366e-01 -5.71274638e-01 5.75426877e-01 -1.32103190e-01
-6.40281916e-01 1.07581580e+00 5.13856374e-02 9.05042589e-01
-6.49481773e-01 1.99924067e-01 2.65122563e-01 6.07839972e-02
5.35533011e-01 4.64729279e-01 -3.65842760e-01 2.26402074e-01
-8.23026896e-01 3.93930644e-01 8.00103366e-01 3.70811820e-02
3.58803511e-01 -1.59159489e-02 -4.00154233e-01 6.17968321e-01
7.81971455e-01 7.13187218e-01 2.37624407e-01 -8.78803730e-01
-1.94392443e-01 6.17295980e-01 -6.90820143e-02 -1.04389596e+00
-3.07099879e-01 -2.92688400e-01 -5.66886246e-01 6.29514456e-01
3.68611544e-01 -2.64091522e-01 -8.02023530e-01 1.50043261e+00
4.04828161e-01 -1.23124003e-01 4.79815435e-03 4.31797594e-01
7.53152847e-01 1.51792586e-01 7.29401112e-01 -4.03897732e-01
1.35687900e+00 5.08801006e-02 -8.74172747e-01 1.45449489e-01
5.14253139e-01 -5.19220531e-01 6.07007802e-01 9.16410983e-01
-8.98893297e-01 -1.82604671e-01 -1.03113151e+00 3.76956820e-01
-6.83443308e-01 -3.36770937e-02 8.08800876e-01 1.60034251e+00
-4.07011420e-01 5.40347636e-01 -1.95512757e-01 -3.14000368e-01
5.23135841e-01 3.85458410e-01 -2.11290523e-01 1.50942996e-01
-1.39944649e+00 1.29747605e+00 3.97393405e-01 -1.30259842e-01
-6.22729123e-01 -6.34028792e-01 -7.09932208e-01 -2.00551197e-01
1.66929364e-01 -2.52987266e-01 1.00994849e+00 -1.24738741e+00
-1.48500562e+00 1.10455167e+00 7.94902369e-02 -5.78087330e-01
6.44944906e-01 -1.63723201e-01 -6.52460217e-01 -1.80188552e-01
2.50940397e-02 3.84588242e-01 5.48830211e-01 -1.16755939e+00
-5.47272086e-01 -6.37401283e-01 -1.22703828e-01 -4.50566024e-01
-2.67248303e-01 6.14524543e-01 6.44071698e-01 -4.96638268e-01
-2.66181201e-01 -8.30188572e-01 -1.52562425e-01 -4.30198193e-01
-1.07280165e-01 -5.78742743e-01 2.77104914e-01 -3.27745646e-01
1.23180020e+00 -1.71437812e+00 -2.98450351e-01 8.38308454e-01
-2.81303078e-01 3.94651681e-01 2.30953276e-01 1.25235066e-01
-1.79279074e-01 3.38890314e-01 -3.19644451e-01 4.53081995e-01
2.18939558e-01 5.49064809e-03 -8.36589560e-02 4.29395676e-01
2.48441935e-01 4.66478378e-01 -7.12072492e-01 -5.55283546e-01
2.64969766e-01 4.61236268e-01 -4.02755648e-01 -7.23516494e-02
-8.60616490e-02 1.51927739e-01 -5.52207708e-01 6.09419107e-01
9.24982369e-01 6.10467911e-01 3.00513297e-01 -1.94220357e-02
-2.39739224e-01 2.93119103e-01 -1.15176558e+00 9.05927956e-01
-1.05277397e-01 1.63517416e-01 -3.40676486e-01 -1.11499643e+00
1.06357980e+00 3.15925777e-01 3.58849287e-01 -7.23827660e-01
5.38099766e-01 5.15971065e-01 4.47511792e-01 -4.76932973e-01
-1.03670709e-01 -5.98002195e-01 -4.51354347e-02 3.51928532e-01
7.12734535e-02 -1.91029802e-01 -2.17320025e-02 -3.95853549e-01
7.07549989e-01 -5.31120151e-02 7.09767222e-01 -7.73951054e-01
9.73112404e-01 -1.51135266e-01 6.34432912e-01 5.91545045e-01
-3.38297635e-01 -1.83836758e-01 7.62565136e-01 -3.59877020e-01
-4.16618884e-01 -8.08289587e-01 -6.67141497e-01 7.99266815e-01
-5.58025181e-01 1.39272153e-01 -1.15903044e+00 -1.04868531e+00
3.81213158e-01 1.11231458e+00 -9.47461188e-01 -2.46184736e-01
-8.12316090e-02 -1.06406879e+00 4.98574942e-01 7.60972127e-02
4.54860628e-01 -1.10633898e+00 -7.77812898e-01 -9.27691609e-02
5.53208292e-01 -5.32598019e-01 6.24214649e-01 5.01618832e-02
-8.46778572e-01 -1.10572839e+00 -5.95295429e-01 3.05506662e-02
4.24144000e-01 -5.14512539e-01 9.87133741e-01 8.81065503e-02
-1.99868530e-01 4.31590259e-01 -3.41562003e-01 -1.57021809e+00
-8.34069669e-01 -1.06889732e-01 1.09165519e-01 8.14056490e-03
1.22555780e+00 -4.16367173e-01 -4.22005177e-01 3.97086591e-01
-1.02207422e+00 -7.09880352e-01 8.13342631e-01 2.94250667e-01
4.28060675e-03 -9.93355811e-02 9.92549241e-01 -1.32316315e+00
8.33021700e-01 -5.78103244e-01 -9.02063191e-01 3.09004754e-01
-1.15855145e+00 2.48066723e-01 -2.86811858e-01 -1.47155702e-01
-1.22959793e+00 -2.45952085e-01 -2.04204261e-01 4.29082721e-01
-5.54931939e-01 3.88597012e-01 -3.73724282e-01 -4.01428044e-01
8.90833676e-01 -7.03094959e-01 9.75581631e-02 -3.30247462e-01
5.51845282e-02 8.00559521e-01 -1.15854360e-01 -6.78306103e-01
6.27702892e-01 2.37540275e-01 5.38457274e-01 -8.26136827e-01
-5.81929266e-01 -7.31912255e-02 -4.21614796e-01 -6.29239082e-01
1.01707530e+00 -2.86465734e-01 -9.61502850e-01 1.91277146e-01
-1.07141697e+00 5.64358644e-02 -7.57449269e-02 9.32463229e-01
-3.39086473e-01 1.57130823e-01 9.73686278e-02 -1.45316732e+00
-3.32646251e-01 -8.75108778e-01 6.53126895e-01 1.91566318e-01
-5.61771810e-01 -7.14135468e-01 2.46853650e-01 4.64104176e-01
3.54321897e-01 2.18563646e-01 1.02038503e+00 -9.29613054e-01
6.19812757e-02 -1.11982778e-01 2.02688593e-02 6.51944041e-01
-3.00450660e-02 1.79272443e-01 -1.13978231e+00 3.47624272e-01
1.36197373e-01 6.35089725e-02 8.30409169e-01 4.73117799e-01
9.50681090e-01 -9.32082534e-02 -1.89251557e-01 -2.50587553e-01
1.35467434e+00 6.18993461e-01 1.03137004e+00 3.17442000e-01
9.37966630e-02 1.45116377e+00 1.06276834e+00 2.34521225e-01
-1.70390666e-01 5.58458626e-01 2.54274309e-01 2.84597218e-01
2.46670038e-01 1.64702877e-01 4.20185477e-01 2.21029222e-01
-4.18580621e-01 9.98034999e-02 -1.11729074e+00 2.21198156e-01
-1.57224643e+00 -1.02099741e+00 -3.29838693e-01 2.49216413e+00
6.27906322e-01 4.41614062e-01 2.17969656e-01 4.07690823e-01
5.58544576e-01 -2.82921374e-01 -6.27861777e-03 -1.24029946e+00
9.39952806e-02 5.90009570e-01 4.83508527e-01 5.35623968e-01
-1.08204842e+00 5.99252284e-01 7.38109016e+00 6.62875116e-01
-9.73832965e-01 1.90004423e-01 7.26492167e-01 5.86938262e-02
-5.05247772e-01 -9.29287970e-02 -7.51840889e-01 4.65690494e-01
1.25191474e+00 -2.16198176e-01 -1.93560317e-01 7.48517275e-01
2.20806435e-01 -5.80665112e-01 -1.01941395e+00 5.50313652e-01
1.67282164e-01 -9.11675632e-01 1.56654954e-01 4.90127385e-01
6.01049840e-01 -3.85461718e-01 3.18660773e-02 6.67499751e-02
2.18770847e-01 -1.19794869e+00 4.86428529e-01 8.31962109e-01
2.94964373e-01 -9.39661682e-01 1.05178976e+00 -5.65137193e-02
-6.32703081e-02 -1.05488203e-01 -3.72284770e-01 -2.30532214e-01
-2.97680020e-01 1.24684668e+00 -5.98836124e-01 4.37867045e-01
5.84217310e-01 9.02696922e-02 -4.11642015e-01 7.74888277e-01
-4.18450534e-01 7.94018745e-01 -1.50442898e-01 -4.10752594e-01
1.84834413e-02 -4.03420091e-01 5.16661227e-01 1.30204058e+00
3.12141001e-01 1.26910239e-01 -6.73083901e-01 9.88283634e-01
2.89107233e-01 4.94368792e-01 -9.82892454e-01 -1.11284137e-01
1.46492049e-01 1.05416858e+00 -6.64039373e-01 -1.29836977e-01
-3.96581650e-01 2.76310593e-01 -4.11584854e-01 -7.28307813e-02
-4.61753249e-01 -3.01072504e-02 6.17624044e-01 1.56627774e-01
-3.22380662e-01 2.42976367e-01 -5.01096547e-01 -6.12935185e-01
-1.33316800e-01 -1.08693480e+00 6.10886931e-01 -3.16675752e-01
-1.30522764e+00 2.07090646e-01 7.66083181e-01 -7.62811959e-01
-4.26522195e-01 -1.09913170e+00 -4.44030166e-01 9.69592869e-01
-1.36044407e+00 -9.51440036e-01 5.48225567e-02 2.39901226e-02
-1.40859619e-01 -5.32805026e-01 8.58443975e-01 2.09869221e-01
-4.07645375e-01 4.94631052e-01 -3.40434819e-01 -1.45777225e-01
7.73804367e-01 -1.15831411e+00 -1.47808701e-01 5.90718329e-01
-1.80132747e-01 4.30390298e-01 1.01955521e+00 -8.86223197e-01
-8.59322250e-01 -4.46522713e-01 1.11581635e+00 -4.93945122e-01
3.11701864e-01 -3.12527046e-02 -6.94894731e-01 2.14007720e-01
2.68394500e-01 -4.57150787e-01 1.22591257e+00 3.31821412e-01
-3.51936460e-01 -2.72712708e-01 -1.74366975e+00 3.42722178e-01
5.81346333e-01 -1.92801163e-01 -9.69506085e-01 -1.29866228e-02
-3.59451510e-02 5.22271156e-01 -9.09879923e-01 6.84275985e-01
8.88192177e-01 -1.21550298e+00 6.76182389e-01 -7.17297792e-01
3.46335471e-01 -1.04001895e-01 -1.38200279e-02 -1.13265634e+00
-7.80898407e-02 -3.27842474e-01 3.87388796e-01 1.37531567e+00
4.53046232e-01 -1.13711703e+00 7.29067326e-01 9.87177074e-01
5.96696913e-01 -4.32975858e-01 -1.10175872e+00 -7.81130016e-01
5.73499799e-01 -6.89825177e-01 7.01241434e-01 1.03591752e+00
-5.76189794e-02 -1.35336772e-01 -2.51489393e-02 1.97465103e-02
8.92978966e-01 -5.15015304e-01 4.91989225e-01 -1.90352607e+00
2.14566708e-01 -6.35968685e-01 -9.04729664e-01 5.87852597e-01
3.83652925e-01 -6.24231815e-01 -2.99175709e-01 -1.19468677e+00
1.78440511e-01 -1.32648304e-01 -7.43076563e-01 1.40109777e-01
9.29055288e-02 -4.66008997e-03 -2.34850729e-03 -5.08024156e-01
3.00716370e-01 1.10631488e-01 5.38822889e-01 -7.36601204e-02
2.35597134e-01 6.94404007e-04 -9.49754596e-01 1.00925446e+00
1.02254319e+00 -8.74096632e-01 -2.41878390e-01 2.50744283e-01
5.41960597e-01 -6.54347777e-01 4.91773158e-01 -1.24166310e+00
-2.80913979e-01 -6.44160211e-01 4.48441803e-01 -2.57247180e-01
-1.52648419e-01 -1.09260976e+00 1.51603237e-01 7.56016076e-01
-5.08658290e-01 -2.19127998e-01 2.62451857e-01 1.36901483e-01
-9.10989568e-02 -8.45556617e-01 7.20676303e-01 -1.10801868e-01
-2.52436519e-01 -2.38845408e-01 -6.01594090e-01 -3.45805138e-01
1.39641511e+00 -5.70877455e-02 -2.65497148e-01 -1.15930885e-01
-3.57541889e-01 -3.80979627e-02 2.85786372e-02 3.14788431e-01
2.50880003e-01 -1.32755482e+00 -8.30132902e-01 -7.81222805e-02
2.85127610e-01 -9.24500465e-01 9.51415747e-02 7.03254640e-01
-3.77544850e-01 6.72053397e-01 -5.83461642e-01 -8.49801451e-02
-1.50896800e+00 6.37260318e-01 3.40936393e-01 -1.39323160e-01
3.11284214e-01 3.83974999e-01 9.73731950e-02 -1.43547475e-01
1.90962106e-01 -2.10865512e-01 -9.24602926e-01 4.80338722e-01
6.18610919e-01 7.16710627e-01 2.01535016e-01 -5.68668425e-01
-6.15805328e-01 6.58976674e-01 1.54327095e-01 -5.64334691e-01
1.19324934e+00 4.87654299e-01 -5.27952433e-01 5.72755873e-01
7.69553840e-01 2.45229349e-01 -2.04181731e-01 3.42911005e-01
6.21421576e-01 -4.16939795e-01 3.12576979e-01 -1.25021207e+00
-6.57903075e-01 8.44101548e-01 1.13426340e+00 3.16671371e-01
9.71598864e-01 -2.30690718e-01 -1.92016721e-01 3.85105520e-01
3.12684089e-01 -1.57355142e+00 -4.56286609e-01 8.90709683e-02
1.20840728e+00 -1.14648533e+00 2.55818516e-01 -7.36115575e-01
-4.47718024e-01 1.05575430e+00 3.24858218e-01 -1.22927641e-02
1.03101754e+00 7.62804002e-02 1.56139314e-01 -3.27494711e-01
-4.12255883e-01 -4.40226346e-01 7.00133145e-01 7.25416601e-01
9.30179477e-01 4.16159332e-01 -1.48755658e+00 7.52768517e-01
1.36646077e-01 5.20363808e-01 2.89123863e-01 7.80779064e-01
-4.10754830e-01 -1.76383865e+00 -7.40607858e-01 6.70282423e-01
-8.39462101e-01 2.89389461e-01 -1.03961384e+00 7.64264166e-01
4.98509735e-01 1.10825372e+00 -2.17358440e-01 -3.27933133e-01
4.43165898e-01 4.35148627e-01 5.22679806e-01 -4.84804690e-01
-9.22208488e-01 -3.42056513e-01 5.07220507e-01 -4.35518205e-01
-9.65142667e-01 -9.60354567e-01 -8.71546447e-01 -1.20644838e-01
-1.15581207e-01 3.97342563e-01 1.12947309e+00 1.07046366e+00
8.69930163e-02 4.57790762e-01 3.28797340e-01 -3.16569597e-01
-4.65178519e-01 -1.03322339e+00 -5.02133012e-01 1.70905367e-01
-1.48651987e-01 -1.13813806e+00 -4.24010485e-01 -5.03227592e-01] | [8.826452255249023, 4.900684356689453] |
a75427e9-715d-47c1-9004-0073409f7abf | a-grounded-unsupervised-universal-part-of | 1904.05426 | null | http://arxiv.org/abs/1904.05426v1 | http://arxiv.org/pdf/1904.05426v1.pdf | A Grounded Unsupervised Universal Part-of-Speech Tagger for Low-Resource Languages | Unsupervised part of speech (POS) tagging is often framed as a clustering
problem, but practical taggers need to \textit{ground} their clusters as well.
Grounding generally requires reference labeled data, a luxury a low-resource
language might not have. In this work, we describe an approach for low-resource
unsupervised POS tagging that yields fully grounded output and requires no
labeled training data. We find the classic method of Brown et al. (1992)
clusters well in our use case and employ a decipherment-based approach to
grounding. This approach presumes a sequence of cluster IDs is a `ciphertext'
and seeks a POS tag-to-cluster ID mapping that will reveal the POS sequence. We
show intrinsically that, despite the difficulty of the task, we obtain
reasonable performance across a variety of languages. We also show
extrinsically that incorporating our POS tagger into a name tagger leads to
state-of-the-art tagging performance in Sinhalese and Kinyarwanda, two
languages with nearly no labeled POS data available. We further demonstrate our
tagger's utility by incorporating it into a true `zero-resource' variant of the
Malopa (Ammar et al., 2016) dependency parser model that removes the current
reliance on multilingual resources and gold POS tags for new languages.
Experiments show that including our tagger makes up much of the accuracy lost
when gold POS tags are unavailable. | ['Ying Lin', 'Heng Ji', 'Ronald Cardenas', 'Jonathan May'] | 2019-04-10 | a-grounded-unsupervised-universal-part-of-1 | https://aclanthology.org/N19-1252 | https://aclanthology.org/N19-1252.pdf | naacl-2019-6 | ['decipherment'] | ['natural-language-processing'] | [-2.14926094e-01 3.40282440e-01 -1.44775882e-01 -4.27805185e-01
-1.33418620e+00 -1.16089535e+00 3.97658050e-01 2.39334688e-01
-5.03907025e-01 7.95204699e-01 3.13842922e-01 -8.08338821e-01
2.99875945e-01 -4.25393581e-01 -4.73648638e-01 -5.74513137e-01
-7.02423975e-02 8.05818975e-01 4.89126682e-01 -3.34721953e-01
-8.94583613e-02 1.61868349e-01 -1.04668581e+00 3.93410057e-01
6.29575074e-01 1.96114942e-01 5.37333906e-01 4.11143720e-01
-3.65858734e-01 6.50366902e-01 -6.85798526e-01 -7.03020334e-01
2.00658694e-01 -2.61197835e-01 -1.24643505e+00 -1.65043890e-01
1.45273820e-01 4.31756020e-01 4.61848676e-02 1.02855170e+00
4.15323287e-01 -8.51638094e-02 3.71394485e-01 -9.41329300e-01
-6.16560519e-01 1.26437068e+00 -2.69349784e-01 2.50579506e-01
2.11955905e-01 -3.48149836e-01 1.33135653e+00 -7.29215145e-01
8.83094370e-01 7.51616001e-01 1.09870470e+00 5.44979334e-01
-1.20516098e+00 -4.60229188e-01 -4.83618230e-02 -3.40646237e-01
-1.53914261e+00 -4.73031908e-01 3.02122563e-01 -4.68390018e-01
1.20571828e+00 2.68641170e-02 2.99684286e-01 7.20563829e-01
-1.67027995e-01 5.25565684e-01 1.20449317e+00 -1.00926018e+00
2.40131319e-01 1.25676349e-01 1.42587438e-01 6.75252199e-01
4.39327091e-01 -6.98359981e-02 -2.71583825e-01 -2.92437613e-01
4.76389140e-01 -1.54265285e-01 1.74075618e-01 -1.98181212e-01
-1.24972224e+00 7.19632089e-01 1.32086053e-01 9.07457709e-01
-1.18714556e-01 -9.92375687e-02 3.17680240e-01 3.80017497e-02
7.11549163e-01 5.76602757e-01 -9.48878109e-01 -3.50917578e-01
-1.10578060e+00 -3.04665148e-01 1.10412598e+00 1.22664583e+00
9.18107390e-01 -2.40515068e-01 2.11793736e-01 9.78638113e-01
2.83408821e-01 5.34371138e-01 5.05950928e-01 -8.33424032e-01
5.29950738e-01 2.38073364e-01 3.50255340e-01 -3.99041921e-01
-4.51228708e-01 -2.24988639e-01 -1.65808156e-01 -1.69495419e-01
6.27295852e-01 -4.33262914e-01 -9.77458715e-01 1.64847612e+00
2.75822341e-01 -4.22244370e-02 3.82137716e-01 5.00654936e-01
4.92249995e-01 3.72081518e-01 4.88401681e-01 -1.80040359e-01
1.72483087e+00 -7.20694005e-01 -4.64488536e-01 -4.52880710e-01
1.20137024e+00 -1.01718748e+00 8.65632713e-01 5.64546101e-02
-9.14952338e-01 -1.95577919e-01 -8.56254995e-01 4.58823927e-02
-5.01170218e-01 -2.70711500e-02 1.08057821e+00 8.37843895e-01
-1.43677485e+00 6.14387333e-01 -1.14568508e+00 -8.94093335e-01
2.83219703e-02 5.24392486e-01 -6.62847400e-01 -3.18759568e-02
-1.09355640e+00 1.04812217e+00 6.17863715e-01 -1.62887484e-01
-9.85504687e-02 -2.01843843e-01 -7.14680254e-01 -1.16533339e-01
1.66087195e-01 -1.08337417e-01 1.55147493e+00 -9.72698808e-01
-1.05668819e+00 1.32077098e+00 -1.21218182e-01 -2.32507139e-01
-1.99496195e-01 -2.66920682e-02 -6.64842486e-01 3.73672955e-02
5.67431569e-01 6.89868629e-01 1.58742353e-01 -1.16743052e+00
-1.00555027e+00 -1.05429173e-01 -4.12894152e-02 -4.44620987e-03
-1.42103523e-01 7.27328718e-01 -3.81613970e-01 -5.73497951e-01
2.69682318e-01 -1.16314805e+00 -2.98686951e-01 -1.00289106e+00
-1.51829123e-01 -2.56992042e-01 1.14541203e-01 -8.44842672e-01
1.23243129e+00 -2.09054947e+00 -4.33188617e-01 2.92809993e-01
-5.46572804e-02 3.58182341e-01 -9.56541747e-02 7.12843239e-01
-1.04694955e-01 5.46430469e-01 -9.66699049e-02 -3.29272807e-01
1.38488144e-01 8.04825604e-01 -1.01560950e-01 4.43665594e-01
2.59571165e-01 9.51350152e-01 -1.17233157e+00 -7.35424578e-01
-2.38240406e-01 2.59920985e-01 -4.47932035e-01 -1.47894472e-02
3.68291214e-02 1.86544150e-01 -7.43809268e-02 7.35697925e-01
3.87692720e-01 -1.06121637e-01 1.02293050e+00 3.77852321e-01
-5.04945338e-01 7.85058737e-01 -1.07191765e+00 1.80364001e+00
-2.64187843e-01 2.72869051e-01 1.24465421e-01 -8.19154620e-01
8.96591544e-01 5.29945910e-01 4.53482091e-01 -1.84253827e-01
-7.73048773e-02 7.12573886e-01 2.74376839e-01 -2.75166571e-01
7.14158893e-01 -6.75397217e-01 -6.56223238e-01 4.79231834e-01
3.98394167e-01 2.03142539e-01 1.22918591e-01 7.95894563e-02
1.46953404e+00 4.54051435e-01 5.19476533e-01 -5.85639596e-01
7.36268284e-03 7.35180855e-01 8.72946918e-01 6.79866195e-01
-2.48576760e-01 4.57993507e-01 2.44016826e-01 -1.64520025e-01
-1.32647097e+00 -9.07858729e-01 -2.66909510e-01 1.53807771e+00
-4.48642910e-01 -4.24193382e-01 -7.73975134e-01 -1.00779033e+00
-3.99155140e-01 4.53837693e-01 -4.35323447e-01 5.08769035e-01
-8.64874363e-01 -7.93431342e-01 9.76002693e-01 5.82808197e-01
-1.63382679e-01 -1.24135125e+00 -1.46023989e-01 5.69075644e-01
-2.44212717e-01 -1.13545024e+00 -2.15805888e-01 9.25898910e-01
-6.01894617e-01 -7.63955176e-01 -4.53988194e-01 -1.41269708e+00
6.38912678e-01 7.91467652e-02 1.33898604e+00 1.67249620e-01
3.26355934e-01 -1.29662491e-02 -9.71291661e-01 -2.45907605e-01
-7.32666075e-01 5.33589184e-01 1.51218567e-02 -6.32011831e-01
8.91001701e-01 -5.08301914e-01 -3.09524890e-02 1.32218331e-01
-6.26340151e-01 -3.37049216e-01 6.48442566e-01 8.34381223e-01
3.93049777e-01 -1.65813193e-01 4.62559432e-01 -1.38731587e+00
3.83892953e-02 -6.70518279e-01 -4.24364239e-01 1.68629080e-01
-3.03375125e-01 1.65088370e-01 5.12154162e-01 -1.99010193e-01
-9.17810917e-01 2.93014795e-01 -4.93036807e-01 1.71171233e-01
-3.82592469e-01 5.84249377e-01 -3.59418631e-01 1.61131233e-01
6.84284031e-01 -3.04995835e-01 -3.39573354e-01 -8.23321164e-01
4.53564048e-01 1.05779338e+00 7.66705215e-01 -8.77842009e-01
7.50611603e-01 2.58435577e-01 -1.94121614e-01 -3.91947210e-01
-1.04199517e+00 -1.05532348e+00 -1.19508171e+00 3.26757312e-01
9.15127456e-01 -1.11438060e+00 -2.36559898e-01 -1.77658349e-01
-1.16494405e+00 -6.39886260e-01 -3.08856219e-01 5.09019017e-01
-3.67979079e-01 4.07707930e-01 -7.20252931e-01 -5.75393260e-01
-8.13539419e-03 -5.90537667e-01 9.13977742e-01 -1.95884600e-01
-4.33910936e-01 -1.26096058e+00 3.82682204e-01 6.13947064e-02
-7.37850815e-02 3.60993966e-02 8.20572972e-01 -1.50055647e+00
-6.33404255e-02 -7.68717900e-02 5.20587526e-02 2.24198282e-01
1.29307851e-01 -2.67563134e-01 -9.49574769e-01 -1.96868092e-01
-3.67457867e-01 -1.43505745e-02 3.16394150e-01 -4.42530103e-02
8.22615251e-02 -3.70215565e-01 -4.05830413e-01 3.27782273e-01
1.53749371e+00 3.01107258e-01 5.01380980e-01 5.22087991e-01
7.99842119e-01 7.25921035e-01 7.74316967e-01 3.26454304e-02
6.07890248e-01 5.72808743e-01 -2.24982083e-01 -1.70746505e-01
-8.06913823e-02 -4.25454587e-01 4.93008912e-01 1.52862215e+00
-7.14407191e-02 -1.69310778e-01 -1.46273232e+00 9.25333858e-01
-1.86738253e+00 -9.84872937e-01 -2.74055690e-01 2.06070352e+00
1.23249984e+00 7.78904185e-02 1.50125951e-01 -5.37341908e-02
9.51759458e-01 -3.34184110e-01 3.45621645e-01 -5.73360682e-01
-3.01863737e-02 7.62548745e-01 7.94077575e-01 6.89498544e-01
-1.27251720e+00 1.65782011e+00 6.40744448e+00 6.35387599e-01
-8.20208728e-01 6.78241014e-01 8.28941762e-02 5.22042096e-01
-2.15076774e-01 5.53349197e-01 -9.92801964e-01 4.36536252e-01
1.35350013e+00 3.25085640e-01 3.24723303e-01 8.35212052e-01
-1.34270474e-01 -3.30301374e-02 -1.02686715e+00 5.71153283e-01
-1.13630578e-01 -1.00162566e+00 -6.39322937e-01 3.05163890e-01
6.59497976e-01 5.57062745e-01 -5.67209423e-01 3.70907933e-01
1.37566197e+00 -7.88569629e-01 9.70215142e-01 6.93687648e-02
9.97202873e-01 -6.35563910e-01 1.04673707e+00 3.26316297e-01
-1.11054039e+00 2.33057007e-01 -5.51357090e-01 -2.11554274e-01
3.01513165e-01 4.10712272e-01 -1.16675341e+00 4.81516778e-01
5.01042545e-01 3.71584624e-01 -5.66006780e-01 8.10719311e-01
-5.78336418e-01 1.09473348e+00 -3.85327548e-01 -5.51983938e-02
4.39656168e-01 5.43391146e-02 1.56274602e-01 1.69520390e+00
4.45578754e-01 2.36283660e-01 5.29303074e-01 5.14625944e-03
4.91925105e-02 1.99016243e-01 -5.90793431e-01 -9.65859592e-02
8.71963203e-01 1.33024943e+00 -1.22221220e+00 -4.37716126e-01
-5.32178521e-01 8.58434319e-01 5.49051344e-01 1.04477942e-01
-2.79790193e-01 -3.13603342e-01 5.26755929e-01 1.30027086e-01
7.71597326e-01 -3.93342435e-01 -2.16351330e-01 -1.10767627e+00
-2.50363141e-01 -7.09664345e-01 5.48949957e-01 -8.01865697e-01
-1.33097494e+00 6.26993477e-01 -7.03634173e-02 -9.77923334e-01
-6.40280068e-01 -6.65936351e-01 -1.25744015e-01 1.02804959e+00
-1.21075785e+00 -1.53027809e+00 4.02309716e-01 2.49328271e-01
-2.82205734e-02 -3.34836654e-02 1.19734180e+00 3.57037604e-01
-1.96499273e-01 6.06596231e-01 1.43108964e-01 6.73650801e-01
7.89840698e-01 -1.51984310e+00 6.81780279e-01 1.09871626e+00
6.38441920e-01 9.69389796e-01 6.59991503e-01 -8.46181393e-01
-9.38301861e-01 -1.01565385e+00 1.89024580e+00 -9.49474216e-01
1.13613522e+00 -6.57129228e-01 -7.48918831e-01 1.11705935e+00
2.67928809e-01 1.47047434e-02 1.15080822e+00 6.04221344e-01
-3.74335796e-01 2.88590521e-01 -1.01910603e+00 3.23715359e-01
1.27281070e+00 -6.04454756e-01 -9.60664511e-01 3.48465592e-01
6.83606446e-01 -8.26018453e-02 -1.16663396e+00 -3.80662791e-02
2.99644500e-01 -5.73334396e-01 4.94884491e-01 -6.96174026e-01
-1.07196562e-01 -5.24206400e-01 -4.26462948e-01 -1.32231915e+00
-6.43229723e-01 -9.06186223e-01 6.56131625e-01 1.87597883e+00
7.06984639e-01 -5.14420211e-01 6.19209051e-01 4.49447483e-01
-6.54577076e-01 -1.29851336e-02 -9.20021296e-01 -9.80388343e-01
4.13521796e-01 -5.58863401e-01 6.83383703e-01 1.43831885e+00
6.57886207e-01 4.97103989e-01 1.18337654e-01 4.00702804e-01
2.35361978e-01 -2.63579488e-01 3.95083487e-01 -1.44770563e+00
-4.25194025e-01 2.83177178e-02 -5.04754543e-01 -5.29561579e-01
3.99895489e-01 -1.33826673e+00 6.43253982e-01 -1.60567176e+00
1.11394718e-01 -1.30465972e+00 -4.36080724e-01 1.28057170e+00
-2.84232013e-02 4.88672644e-01 2.59596467e-01 6.89635932e-01
-7.54091024e-01 -5.48675597e-01 3.44015718e-01 3.67623925e-01
-1.41573772e-01 -3.80220145e-01 -1.01580179e+00 7.57528067e-01
8.68174911e-01 -1.06313884e+00 1.29864514e-01 -4.93080050e-01
4.90695596e-01 -2.16964453e-01 3.07694133e-02 -8.51930916e-01
1.59938425e-01 -1.52035087e-01 1.87382214e-02 -1.50387987e-01
-1.10712081e-01 -6.33568645e-01 2.65316874e-01 1.50659278e-01
1.34499565e-01 3.51316899e-01 1.21446654e-01 -1.90763827e-03
-7.44444206e-02 -5.85111260e-01 5.02637506e-01 -3.75548244e-01
-9.17521179e-01 -9.41962376e-02 -4.01932091e-01 3.65876824e-01
6.05584741e-01 -9.04336274e-02 -2.95090914e-01 2.38924190e-01
-7.37164080e-01 -2.73229241e-01 9.74067628e-01 6.55001253e-02
-3.53053808e-01 -1.16498721e+00 -5.13858557e-01 1.29348310e-02
1.72262534e-01 -2.71338224e-01 -3.95039320e-01 5.88546097e-01
-5.67042530e-01 4.90711659e-01 -7.60423094e-02 -3.02408397e-01
-9.69932556e-01 6.77581608e-01 -2.62458682e-01 -4.24819171e-01
-4.15380269e-01 7.18850911e-01 -2.70938843e-01 -6.56724751e-01
-3.85927022e-01 -2.15420693e-01 -1.51398018e-01 1.62247226e-01
7.58477598e-02 -2.25119606e-01 2.81833440e-01 -1.01249635e+00
-6.44471705e-01 1.44009620e-01 1.24884799e-01 -6.02734029e-01
1.50242198e+00 -1.32473662e-01 -2.03815177e-01 4.30379719e-01
7.81669319e-01 6.95690572e-01 -9.18806672e-01 -4.31603454e-02
7.04393864e-01 -2.55142059e-02 -3.41719329e-01 -8.38503838e-01
-3.75919580e-01 6.93420351e-01 9.90170613e-02 3.77804309e-01
5.33805311e-01 6.03683412e-01 7.20080674e-01 2.57087797e-01
8.14359963e-01 -1.19382632e+00 -5.50016403e-01 9.25388098e-01
-8.46615657e-02 -9.29405212e-01 -2.37634853e-01 -3.79681438e-01
-7.07859516e-01 5.77234805e-01 3.36905234e-02 1.00645415e-01
5.29535294e-01 6.33683264e-01 4.47825164e-01 -2.12341473e-01
-4.68148082e-01 -8.39200079e-01 -1.23328939e-01 1.05324364e+00
7.30591416e-01 4.65787917e-01 -4.85911041e-01 8.77350390e-01
-6.71037138e-01 -3.28467935e-01 3.56423616e-01 1.28037941e+00
-7.18855798e-01 -1.79902482e+00 -3.51170450e-01 6.21297657e-02
-1.09254885e+00 -8.15422535e-01 -2.95904011e-01 9.96276617e-01
6.36455357e-01 1.05228221e+00 2.47576997e-01 -3.50094676e-01
5.26849553e-02 5.73439956e-01 3.49221915e-01 -1.37997842e+00
-9.08724546e-01 3.03764969e-01 6.39472127e-01 9.45093408e-02
-7.44701087e-01 -9.87686813e-01 -1.51600838e+00 -2.64292806e-01
-4.59462076e-01 7.27063179e-01 6.86606646e-01 1.08907986e+00
2.92457432e-01 -8.17925632e-02 3.28335077e-01 -5.52722633e-01
-2.05233425e-01 -1.05643713e+00 -8.06486666e-01 3.46639246e-01
-3.30158085e-01 -5.55133939e-01 -1.99728347e-02 3.91791791e-01] | [10.36866569519043, 9.927753448486328] |
1793d958-17c6-41e5-89b2-465e634419d3 | dialogpt-large-scale-generative-pre-training-1 | null | null | https://aclanthology.org/2020.acl-demos.30 | https://aclanthology.org/2020.acl-demos.30.pdf | DIALOGPT : Large-Scale Generative Pre-training for Conversational Response Generation | We present a large, tunable neural conversational response generation model, DIALOGPT (dialogue generative pre-trained transformer). Trained on 147M conversation-like exchanges extracted from Reddit comment chains over a period spanning from 2005 through 2017, DialoGPT extends the Hugging Face PyTorch transformer to attain a performance close to human both in terms of automatic and human evaluation in single-turn dialogue settings. We show that conversational systems that leverage DialoGPT generate more relevant, contentful and context-consistent responses than strong baseline systems. The pre-trained model and training pipeline are publicly released to facilitate research into neural response generation and the development of more intelligent open-domain dialogue systems. | ['Yen-Chun Chen', 'Xiang Gao', 'Chris Brockett', 'Yizhe Zhang', 'Jingjing Liu', 'Jianfeng Gao', 'Siqi Sun', 'Michel Galley', 'Bill Dolan'] | 2020-07-01 | null | null | null | acl-2020-6 | ['conversational-response-generation'] | ['natural-language-processing'] | [ 2.91129529e-01 6.93006516e-01 4.58611213e-02 -7.45345950e-01
-1.20483184e+00 -9.00204718e-01 1.15469193e+00 -3.17532033e-01
-1.49687812e-01 1.24689209e+00 1.02182913e+00 -3.47370058e-01
5.00485718e-01 -6.25597537e-01 -1.97869927e-01 -1.26686454e-01
2.65579551e-01 1.16419291e+00 -2.35700428e-01 -1.00033021e+00
6.02019727e-02 -3.35997611e-01 -6.46443546e-01 9.83983994e-01
8.02231133e-01 5.25141478e-01 -2.18212217e-01 1.22762573e+00
-1.72532108e-02 1.19630241e+00 -1.11761677e+00 -9.39836681e-01
5.45719452e-02 -9.50360298e-01 -1.56780088e+00 -1.66820556e-01
2.55665332e-01 -5.25346398e-01 -2.98494458e-01 3.41584116e-01
8.61304939e-01 5.68489015e-01 5.99530578e-01 -8.33827734e-01
-9.82900739e-01 1.37625802e+00 2.24675566e-01 -2.36346256e-02
7.51059353e-01 7.10820317e-01 1.26370239e+00 -8.19434166e-01
9.47206795e-01 1.67863846e+00 4.60386962e-01 1.29359090e+00
-1.43994606e+00 -5.14928997e-01 -6.03537671e-02 -4.22020078e-01
-3.84445578e-01 -6.85346305e-01 4.86677647e-01 -2.01691076e-01
1.34320199e+00 4.84579504e-01 2.76245952e-01 2.05731630e+00
-6.33403584e-02 9.18395102e-01 1.00589097e+00 -8.86389613e-02
-1.70888454e-01 1.19638786e-01 1.39222562e-01 1.90895155e-01
-7.36286104e-01 -1.87181324e-01 -6.04927540e-01 -4.64193374e-01
4.24586177e-01 -7.61115134e-01 -3.64037782e-01 5.79223514e-01
-1.25890732e+00 1.25529253e+00 3.92661124e-01 9.19482261e-02
-4.47037965e-01 2.19273511e-02 7.39298940e-01 8.79338026e-01
9.05255735e-01 1.11817610e+00 -4.58227038e-01 -7.41374016e-01
-3.40216458e-01 7.55611598e-01 1.58159959e+00 1.06139648e+00
3.19195092e-01 1.50820524e-01 -8.85154486e-01 1.43418264e+00
-2.04174295e-01 3.68399203e-01 5.63009381e-01 -1.21913624e+00
8.69460464e-01 5.56100607e-01 2.01982439e-01 -4.48897302e-01
-3.91053140e-01 6.01288080e-02 -8.29110920e-01 -4.94454354e-01
5.07842362e-01 -1.09334004e+00 -2.69734800e-01 1.70895135e+00
3.06921136e-02 -7.57899880e-01 4.98709589e-01 9.36997890e-01
1.36256099e+00 9.56376851e-01 -9.47180614e-02 -6.07977174e-02
1.21983731e+00 -1.35073125e+00 -5.02741039e-01 -2.39334881e-01
6.00784421e-01 -7.59436607e-01 1.29263437e+00 1.89778388e-01
-1.51443493e+00 -4.46007043e-01 -4.33368802e-01 -3.52168322e-01
1.03789426e-01 7.59276189e-03 5.57098031e-01 3.46892536e-01
-1.20876253e+00 1.68169469e-01 -7.84105808e-02 -3.66515249e-01
-1.14436418e-01 1.87608134e-02 -1.82874933e-01 3.41005951e-01
-1.67640805e+00 1.05276036e+00 -3.32948193e-02 -3.00883073e-02
-1.00899267e+00 -4.44011480e-01 -6.29262805e-01 -6.61562830e-02
1.96208701e-01 -8.64933372e-01 2.31314421e+00 -9.50294435e-01
-2.19687533e+00 9.29581225e-01 -2.95154899e-02 -7.97342479e-01
7.37976909e-01 -2.72396654e-01 -7.53896534e-02 4.72687222e-02
-1.98121324e-01 9.17491794e-01 3.73349458e-01 -7.86027312e-01
-2.35621661e-01 2.95998931e-01 3.74959797e-01 5.62954605e-01
-7.57442191e-02 3.40939373e-01 1.30814835e-01 -4.28131402e-01
-8.10571015e-01 -1.14757371e+00 -2.68382728e-01 -8.26704860e-01
-6.99494421e-01 -8.38577688e-01 2.77210295e-01 -6.77830100e-01
7.59951353e-01 -1.45266569e+00 2.04342350e-01 -3.35880160e-01
1.23084761e-01 2.55891532e-01 -5.25854230e-01 1.19399905e+00
2.10016668e-01 1.84981674e-02 -1.85778067e-01 -4.77156162e-01
2.00743124e-01 -1.76765114e-01 -7.80501246e-01 -2.86817968e-01
2.70843536e-01 1.21308756e+00 -1.06123829e+00 -8.27267244e-02
-1.15667619e-01 6.65145814e-02 -7.26025045e-01 8.95258963e-01
-8.57089341e-01 8.77008557e-01 -3.55232924e-01 1.03988327e-01
3.40778455e-02 -3.57141048e-01 3.03310722e-01 5.32447219e-01
2.70124115e-02 1.17739463e+00 -9.50373523e-03 1.68018281e+00
-7.03857064e-01 8.48596394e-01 9.88106653e-02 -1.62688658e-01
1.21811664e+00 7.43809998e-01 2.27744859e-02 -5.39936662e-01
1.07344270e-01 -2.88874395e-02 1.67761907e-01 -3.53275329e-01
1.22029293e+00 -2.54936479e-02 -6.37079060e-01 1.32501125e+00
2.47113585e-01 -3.99755508e-01 1.54863268e-01 8.02113235e-01
1.20808637e+00 -2.65206248e-01 1.39871031e-01 -1.57700524e-01
4.94291574e-01 9.22137871e-02 9.86788720e-02 9.38976288e-01
-3.98401022e-02 3.95308018e-01 8.29377294e-01 -3.65102112e-01
-9.68762636e-01 -7.67396867e-01 2.58444875e-01 1.80362570e+00
-5.63055694e-01 -2.48045608e-01 -9.03130889e-01 -7.78622508e-01
-1.52851880e-01 8.53182137e-01 -4.81563866e-01 1.56430714e-02
-8.07757080e-01 -5.28624058e-01 1.06014407e+00 3.46017897e-01
6.28840685e-01 -1.67975640e+00 -1.84902810e-02 3.89349163e-01
-8.79430413e-01 -9.94249761e-01 -8.71074080e-01 -2.86640912e-01
-4.59585398e-01 -7.56674051e-01 -7.80592680e-01 -6.72978461e-01
1.93673491e-01 -1.08687460e-01 1.64902139e+00 3.95657495e-02
1.65244833e-01 2.84179360e-01 -5.19211471e-01 -1.75236702e-01
-1.31400633e+00 6.33826315e-01 -2.65717059e-01 -3.05083007e-01
2.12586224e-01 -3.17964911e-01 -6.27127111e-01 3.59621346e-01
-2.47117192e-01 3.94025028e-01 2.19728410e-01 1.25405562e+00
-5.54829240e-01 -1.35529137e+00 1.24777043e+00 -1.24834740e+00
1.86701190e+00 -6.39285982e-01 -1.91142321e-01 2.21296296e-01
-4.59169030e-01 -7.22104013e-02 7.25990891e-01 -3.81864548e-01
-1.68394113e+00 -5.15632987e-01 -4.28262711e-01 4.08259183e-01
-6.80550262e-02 1.36849061e-01 2.99462020e-01 4.35474396e-01
1.23319757e+00 2.04606608e-01 8.47304687e-02 -3.03905487e-01
8.38025331e-01 1.05938923e+00 7.23581672e-01 -8.91338706e-01
2.81757325e-01 -2.94625551e-01 -9.02547956e-01 -4.75205511e-01
-5.51674902e-01 -1.60783693e-01 -5.79646677e-02 -3.88046414e-01
8.09641838e-01 -8.09665203e-01 -1.22145188e+00 4.53367233e-01
-1.58690369e+00 -1.11134887e+00 -1.68872681e-02 -9.89289060e-02
-6.67381644e-01 -5.85174225e-02 -1.50889087e+00 -8.97654235e-01
-1.15744436e+00 -7.74942398e-01 8.83743525e-01 1.88848585e-01
-1.06286883e+00 -1.09685516e+00 6.63948536e-01 7.88271785e-01
7.48601615e-01 -6.05751574e-02 7.64663279e-01 -1.23966813e+00
-3.88879031e-01 8.84757936e-02 -6.77110925e-02 3.11088055e-01
-8.50115195e-02 -4.81989458e-02 -1.03722334e+00 -9.81850550e-02
-2.26372182e-01 -1.23749959e+00 6.62121654e-01 -9.06187370e-02
4.74147379e-01 -1.00482941e+00 6.27539456e-02 -1.50990183e-03
3.63779068e-01 1.03486188e-01 4.63719755e-01 -2.08103955e-01
3.19120437e-01 1.07578552e+00 3.35049182e-01 5.00256717e-01
6.72907412e-01 5.98185062e-01 -1.85858607e-02 1.43302996e-02
-4.97952402e-02 -5.17808318e-01 5.32917738e-01 1.01465404e+00
4.51982021e-02 -7.94553041e-01 -7.37934470e-01 6.28490627e-01
-1.98336720e+00 -1.26695573e+00 -1.35756016e-01 1.63854825e+00
1.64241052e+00 -5.24736233e-02 5.01563966e-01 -7.54378498e-01
5.96549988e-01 3.50945443e-01 -4.29655224e-01 -1.24480855e+00
-2.11438894e-01 2.48681381e-01 -1.79499224e-01 8.78247917e-01
-5.38311541e-01 1.12219858e+00 6.98314953e+00 2.81379491e-01
-8.48217666e-01 2.32271627e-01 8.30448866e-01 -1.79844648e-01
-5.95141292e-01 2.98526753e-02 -6.64541066e-01 4.68176752e-01
1.37739646e+00 -4.46478814e-01 6.27688885e-01 6.52666509e-01
1.72899663e-01 2.50184178e-01 -1.22507811e+00 3.56890529e-01
5.89247374e-03 -1.47815812e+00 -4.07123007e-02 -7.51067623e-02
9.51201320e-01 1.82318166e-01 -1.13631412e-02 9.77407336e-01
1.49902165e+00 -1.06955922e+00 2.67164916e-01 3.97662044e-01
6.66701734e-01 -4.59752023e-01 5.85743725e-01 4.21022773e-01
-3.56701702e-01 1.22580091e-02 -4.24034297e-02 -4.21315163e-01
5.86931825e-01 1.03327416e-01 -1.84274971e+00 2.45311439e-01
2.53689457e-02 4.41428959e-01 -1.20168999e-01 2.72039652e-01
-5.00296056e-01 9.20760453e-01 1.29697204e-01 -5.07294774e-01
3.69375914e-01 -8.51840451e-02 6.68005526e-01 1.59600627e+00
-3.07935208e-01 3.21193218e-01 1.38148785e-01 1.01397943e+00
-7.58837879e-01 2.11846195e-02 -5.67108154e-01 -2.25471407e-01
6.87572002e-01 1.50475085e+00 1.20657101e-01 -5.34119546e-01
-6.23956919e-02 9.44198191e-01 5.10720253e-01 3.99998933e-01
-4.88567263e-01 -2.11037815e-01 7.42916167e-01 -2.73926288e-01
-3.63953114e-01 3.05094477e-02 -2.53145337e-01 -1.01754153e+00
-3.60089093e-01 -1.47279298e+00 4.28171396e-01 -7.70263016e-01
-1.83930767e+00 1.06417656e+00 -1.01182759e-01 -6.58239245e-01
-1.40239096e+00 -2.02709064e-01 -9.94177043e-01 1.11291552e+00
-7.00733185e-01 -1.10450912e+00 -1.10550866e-01 3.72774124e-01
1.10843074e+00 -2.73795396e-01 1.28419042e+00 -1.90928936e-01
-3.01009417e-01 8.54123950e-01 -3.29315752e-01 5.44559419e-01
1.22651386e+00 -1.33089972e+00 1.29129207e+00 1.74297139e-01
-2.79438615e-01 8.26866210e-01 7.39931941e-01 -5.86535811e-01
-9.14517760e-01 -8.27558339e-01 1.06776404e+00 -9.98953164e-01
5.69184661e-01 -6.92822874e-01 -7.71147013e-01 7.96335101e-01
1.24362111e+00 -9.93302405e-01 7.51901090e-01 5.21948516e-01
-2.35888302e-01 3.71124923e-01 -9.48402703e-01 8.58504772e-01
8.52347314e-01 -8.42619658e-01 -8.01750898e-01 6.75624073e-01
1.14554465e+00 -6.76186085e-01 -9.71810341e-01 5.29459938e-02
5.09028614e-01 -7.42034256e-01 5.22018075e-01 -9.33173776e-01
8.28325987e-01 6.30590618e-01 3.27488989e-01 -1.72978687e+00
-1.80372372e-02 -1.62304866e+00 1.27988219e-01 1.25174654e+00
9.15893435e-01 -7.61728168e-01 4.42264080e-01 9.33703542e-01
-3.87970597e-01 -5.77030420e-01 -6.51991606e-01 -2.06285164e-01
5.38720548e-01 1.68244183e-01 6.68636262e-01 9.90341723e-01
7.93160439e-01 1.30521369e+00 -6.88613534e-01 -8.90238643e-01
-6.39293343e-02 1.80205911e-01 1.43860877e+00 -8.85650873e-01
-5.48539698e-01 -5.33111870e-01 5.86822391e-01 -1.57396865e+00
3.47433239e-01 -8.46679389e-01 3.94113004e-01 -1.44196475e+00
1.81892976e-01 -2.96317548e-01 5.89666545e-01 5.56073189e-01
-2.25690633e-01 2.04824377e-02 1.06688276e-01 1.52015030e-01
-5.27869880e-01 8.33488286e-01 1.42720819e+00 -1.29472718e-01
-2.89680272e-01 2.73931056e-01 -9.83603299e-01 1.95727065e-01
7.52482772e-01 -2.89104789e-01 -4.19489086e-01 -4.88355726e-01
1.61685526e-01 7.38110185e-01 7.24391118e-02 -2.15101480e-01
1.82754219e-01 -2.22718209e-01 -1.55647367e-01 -3.38166445e-01
6.26466155e-01 4.05951083e-01 -3.57638806e-01 1.03819788e-01
-1.48343766e+00 2.35797688e-01 2.05588583e-02 2.29819179e-01
3.43018249e-02 2.83295941e-02 3.85793179e-01 -4.31749016e-01
-2.08649095e-02 -4.51670997e-02 -7.39858806e-01 5.87963343e-01
3.57274979e-01 3.11401159e-01 -1.12252522e+00 -1.41429830e+00
-4.60604548e-01 5.79860687e-01 2.24855378e-01 8.15566719e-01
3.62709641e-01 -9.85688329e-01 -1.37763858e+00 -2.55712152e-01
1.17506713e-01 -1.73241928e-01 2.28458568e-01 4.02868450e-01
-1.76647797e-01 7.39802599e-01 -9.66536552e-02 -1.93161994e-01
-1.01204979e+00 -1.62567645e-01 4.44649160e-01 -6.28692269e-01
-6.21742368e-01 1.14531291e+00 1.74166203e-01 -1.06257153e+00
7.53671825e-02 -7.67603219e-02 -1.25847638e-01 -1.66198105e-01
6.30093813e-01 2.28165165e-01 -1.91605389e-01 -3.13079506e-01
2.33535498e-01 -6.97630584e-01 -3.54859889e-01 -8.17685902e-01
8.62243891e-01 4.17842492e-02 -1.41678631e-01 4.00830656e-01
8.18651497e-01 -1.89893804e-02 -1.19921505e+00 -3.70724410e-01
-1.83651164e-01 -1.41376883e-01 -8.40556324e-01 -1.42413068e+00
-4.16994095e-01 5.67602158e-01 -3.38678420e-01 4.60811198e-01
4.91278946e-01 -1.11841045e-01 1.26442564e+00 8.52134168e-01
2.21218809e-01 -1.04994190e+00 5.33205390e-01 1.18677545e+00
1.39607000e+00 -1.17665398e+00 -5.25396109e-01 9.68528688e-02
-1.37865067e+00 9.18583393e-01 1.10331690e+00 -1.12517901e-01
-1.23789400e-01 1.83381152e-03 4.43588436e-01 -5.14280871e-02
-1.86334407e+00 3.02132964e-01 1.87662497e-01 4.49696064e-01
1.04831135e+00 2.04990387e-01 -3.59416276e-01 6.60313845e-01
-8.59756827e-01 -3.62490207e-01 6.98255718e-01 3.60538840e-01
-2.09398702e-01 -1.20877445e+00 1.24649458e-01 2.69052953e-01
-3.07204455e-01 -2.84114093e-01 -1.44784153e+00 3.93061787e-01
-8.94571960e-01 1.50115621e+00 6.51875231e-03 -5.20900726e-01
3.71111363e-01 3.72870356e-01 1.79996222e-01 -9.31903183e-01
-1.72559917e+00 -3.94425422e-01 1.27979553e+00 -3.10308516e-01
-3.60791124e-02 -3.84633064e-01 -1.06609488e+00 -6.61672354e-01
-1.90624237e-01 4.52678025e-01 2.66266704e-01 7.47030079e-01
4.22733396e-01 2.52932161e-01 8.23974192e-01 -5.81644714e-01
-1.06954825e+00 -1.92222428e+00 3.57754856e-01 5.30349970e-01
1.75882578e-01 1.89873502e-01 -4.81183827e-02 -1.39172748e-01] | [12.687116622924805, 8.200798988342285] |
29e31666-d353-40a3-9b44-df859fe04e9a | document-ranking-with-a-pretrained-sequence | 2003.06713 | null | https://arxiv.org/abs/2003.06713v1 | https://arxiv.org/pdf/2003.06713v1.pdf | Document Ranking with a Pretrained Sequence-to-Sequence Model | This work proposes a novel adaptation of a pretrained sequence-to-sequence model to the task of document ranking. Our approach is fundamentally different from a commonly-adopted classification-based formulation of ranking, based on encoder-only pretrained transformer architectures such as BERT. We show how a sequence-to-sequence model can be trained to generate relevance labels as "target words", and how the underlying logits of these target words can be interpreted as relevance probabilities for ranking. On the popular MS MARCO passage ranking task, experimental results show that our approach is at least on par with previous classification-based models and can surpass them with larger, more-recent models. On the test collection from the TREC 2004 Robust Track, we demonstrate a zero-shot transfer-based approach that outperforms previous state-of-the-art models requiring in-dataset cross-validation. Furthermore, we find that our approach significantly outperforms an encoder-only model in a data-poor regime (i.e., with few training examples). We investigate this observation further by varying target words to probe the model's use of latent knowledge. | ['Rodrigo Nogueira', 'Zhiying Jiang', 'Jimmy Lin'] | 2020-03-14 | null | https://aclanthology.org/2020.findings-emnlp.63 | https://aclanthology.org/2020.findings-emnlp.63.pdf | findings-of-the-association-for-computational | ['ad-hoc-information-retrieval', 'passage-ranking'] | ['natural-language-processing', 'natural-language-processing'] | [ 6.41286850e-01 -8.56159851e-02 -3.49137723e-01 -6.04216456e-01
-1.75444603e+00 -8.22462976e-01 1.45037413e+00 2.22963884e-01
-7.18245029e-01 8.43627274e-01 7.99310565e-01 -3.90526026e-01
-2.76524931e-01 -4.68868166e-01 -7.89803624e-01 -4.21730280e-01
-1.89916998e-01 8.60143483e-01 4.67253476e-01 -8.02002549e-01
7.38373995e-01 -1.44827873e-01 -1.63237488e+00 8.61698270e-01
5.48920631e-01 6.10864401e-01 1.14685334e-01 9.03845906e-01
2.46729672e-01 1.00682378e+00 -6.39491498e-01 -5.76807797e-01
-8.53966773e-02 -3.97375286e-01 -1.21189988e+00 -5.06966054e-01
4.98859823e-01 -3.27430546e-01 -3.12318325e-01 7.19893694e-01
6.40116811e-01 3.11075747e-01 1.07544279e+00 -6.57324195e-01
-9.60954368e-01 7.74531543e-01 1.01886012e-01 4.51920420e-01
5.93451083e-01 -3.84334385e-01 1.56375134e+00 -1.06738544e+00
6.44487739e-01 1.16962051e+00 4.91672635e-01 6.19465828e-01
-1.22766125e+00 -2.47866318e-01 -5.07914387e-02 4.63089645e-01
-1.10505486e+00 -7.10207522e-01 3.51251721e-01 -4.00114506e-01
1.06838083e+00 3.58725429e-01 2.24382713e-01 1.30462527e+00
4.10774350e-02 8.07686567e-01 1.07708406e+00 -9.04461026e-01
3.93440314e-02 3.10130958e-02 3.28087837e-01 3.16218764e-01
-1.38139442e-01 3.13407719e-01 -7.60986984e-01 -2.94008732e-01
2.17808262e-01 -4.53678370e-01 -2.17665568e-01 -3.47392321e-01
-1.22798526e+00 9.32907224e-01 1.87805235e-01 3.58476847e-01
-3.93546782e-02 3.18005264e-01 7.37949669e-01 6.10535204e-01
8.10040772e-01 4.92000222e-01 -6.35613680e-01 -3.49118382e-01
-1.01406586e+00 2.80451685e-01 6.20187402e-01 8.37224126e-01
3.44057441e-01 -1.75326586e-01 -6.48125410e-01 1.08356392e+00
2.56331921e-01 8.02988783e-02 7.53831029e-01 -7.07472801e-01
5.22628427e-01 -1.46367848e-01 4.20106232e-01 -3.16376358e-01
8.38442892e-02 -5.39453566e-01 -1.07852519e-01 -1.55897588e-01
1.71666950e-01 4.00459975e-01 -1.03357077e+00 1.80224419e+00
-4.16087776e-01 5.64396158e-02 2.83876359e-01 7.60794878e-01
5.11701941e-01 7.84868836e-01 2.29024336e-01 -2.13103116e-01
1.23378587e+00 -1.05047560e+00 -3.92340422e-01 -1.79033160e-01
8.78593564e-01 -7.46453047e-01 1.28863609e+00 3.19309413e-01
-1.06950283e+00 -5.64934373e-01 -1.00720751e+00 -2.00143576e-01
-4.11512047e-01 -1.77155118e-02 5.96074224e-01 5.90625644e-01
-1.38821042e+00 9.11117554e-01 -2.54149973e-01 -4.64173019e-01
-1.05261013e-01 1.16103902e-01 -9.61828530e-02 -1.66728035e-01
-1.66015124e+00 1.32716095e+00 5.45141757e-01 -1.56625658e-01
-1.20987523e+00 -4.98680592e-01 -6.59660578e-01 4.84453328e-02
-1.86997876e-02 -5.41755557e-01 1.85398722e+00 -9.22631741e-01
-1.41985536e+00 8.79064143e-01 -1.34232581e-01 -3.65645319e-01
2.40977049e-01 -4.25482035e-01 -4.29713517e-01 1.03148716e-02
2.84093231e-01 5.14633119e-01 4.97615695e-01 -1.22155976e+00
-6.16892040e-01 1.33147851e-01 2.95803040e-01 5.44846117e-01
-5.65083206e-01 5.23852408e-01 -2.73031294e-01 -5.65102875e-01
-5.39922714e-01 -7.96610713e-01 -1.01000899e-02 -5.42806327e-01
-4.90979888e-02 -4.53598201e-01 2.10253168e-02 -5.61827958e-01
1.27645397e+00 -1.78802598e+00 1.52246594e-01 8.82480219e-02
-3.09111685e-01 3.67180169e-01 -4.70348269e-01 9.24997211e-01
-3.30912769e-01 2.54993200e-01 -2.09778044e-02 -2.40813628e-01
1.84656709e-01 -2.54771542e-02 -5.30150473e-01 1.84821010e-01
1.05809487e-01 9.99685645e-01 -1.32894278e+00 -5.10633826e-01
-8.47115368e-02 4.07144964e-01 -4.09537792e-01 1.21778205e-01
-1.72808841e-01 8.52593407e-02 -3.55542421e-01 1.76145509e-01
-1.47056341e-01 -3.57940942e-01 3.26383770e-01 1.78634346e-01
-3.86222755e-03 1.20894921e+00 -4.19739574e-01 1.84294331e+00
-6.43208623e-01 7.56231248e-01 -6.43798292e-01 -1.10058248e+00
7.50966072e-01 6.79696679e-01 -2.29587555e-02 -8.93885076e-01
-1.78600267e-01 4.98685747e-01 8.26048478e-02 -1.21393286e-01
9.58325207e-01 -2.99910963e-01 -2.06027150e-01 4.78151888e-01
1.79682419e-01 1.47529572e-01 4.87349421e-01 5.30393839e-01
1.33826315e+00 3.74092221e-01 1.45157397e-01 -4.50661302e-01
4.49874610e-01 4.98855626e-03 -1.25927627e-01 1.04787052e+00
2.24168107e-01 6.19043171e-01 2.84763694e-01 -8.88940692e-02
-1.30305195e+00 -9.68060195e-01 -3.83710176e-01 1.61189246e+00
-2.20682830e-01 -5.13114929e-01 -4.64216679e-01 -7.73107350e-01
-3.65470827e-01 1.06893075e+00 -6.59042180e-01 -2.99680948e-01
-4.51752305e-01 -6.31828070e-01 5.57307661e-01 5.24458766e-01
-2.57637918e-01 -1.20705473e+00 -8.98649693e-02 4.32844222e-01
-4.31151867e-01 -6.86842680e-01 -3.65470469e-01 3.81562382e-01
-8.77245843e-01 -6.80592418e-01 -9.58349586e-01 -7.45697200e-01
3.88505280e-01 6.76835999e-02 1.63530076e+00 1.05432503e-01
2.00930387e-01 4.82023865e-01 -7.55801260e-01 -1.75003469e-01
-6.93153441e-01 2.90266693e-01 -1.55601770e-01 -3.63672614e-01
5.07155597e-01 -3.52286220e-01 -4.36768800e-01 1.15949214e-01
-9.61599588e-01 -9.84259397e-02 6.75203919e-01 1.23818576e+00
1.26349926e-01 -3.74340326e-01 9.94701087e-01 -1.08460093e+00
1.03711379e+00 -5.74559212e-01 -2.52438873e-01 4.86399472e-01
-8.39361310e-01 3.60887289e-01 5.72584331e-01 -4.21330929e-01
-9.63571191e-01 -3.93541574e-01 -2.40034357e-01 7.42801726e-02
1.90965444e-01 8.26716959e-01 2.69536704e-01 4.23070192e-01
8.35784137e-01 3.61704886e-01 -4.33217824e-01 -4.99006540e-01
5.42506039e-01 8.81828904e-01 4.47299451e-01 -6.98305726e-01
7.57239282e-01 -3.29428837e-02 -3.10782135e-01 -1.66634127e-01
-1.26031649e+00 -7.41992235e-01 -4.88339424e-01 -2.91886628e-01
4.08801973e-01 -1.08825362e+00 -1.31947234e-01 -3.62348855e-02
-1.32247496e+00 -3.63258570e-01 -2.37137750e-01 4.84566659e-01
-8.00424218e-01 2.37089396e-01 -9.77114260e-01 -7.47285843e-01
-2.86370665e-01 -8.12542975e-01 1.29731715e+00 -2.57331938e-01
-3.73095632e-01 -1.14149642e+00 6.12436950e-01 2.19143257e-01
4.35386270e-01 -3.70301545e-01 1.22013855e+00 -1.00753009e+00
-2.60121465e-01 -5.16125083e-01 -1.10917151e-01 4.68324006e-01
-2.80259639e-01 -3.08775991e-01 -1.11785042e+00 -3.52576703e-01
-4.32572991e-01 -8.37295532e-01 1.22753131e+00 -1.99748371e-02
8.49802673e-01 -1.50988117e-01 -2.06558511e-01 -2.13574544e-01
1.36227608e+00 1.08169787e-01 9.89149392e-01 4.90739495e-01
3.02936256e-01 6.74040735e-01 8.65474164e-01 7.43391290e-02
3.71712923e-01 1.00227618e+00 -9.16686258e-04 1.37275979e-01
-1.23365447e-01 -6.01208627e-01 6.67361557e-01 1.28657365e+00
-2.61314064e-01 -6.17282629e-01 -6.39633894e-01 8.27864826e-01
-1.94357800e+00 -1.24691486e+00 6.03675582e-02 2.50683284e+00
1.15778875e+00 3.62575471e-01 -5.68647087e-02 9.96311530e-02
6.19511664e-01 2.35669076e-01 2.06107479e-02 -4.93820250e-01
1.32089913e-01 5.69435835e-01 3.10056418e-01 8.14421296e-01
-1.04095435e+00 8.23226511e-01 7.24644184e+00 1.07372904e+00
-6.36820197e-01 4.29716498e-01 5.07916152e-01 -3.10774352e-02
-6.67245507e-01 1.56695083e-01 -8.77399623e-01 5.26887357e-01
1.61972392e+00 -2.66085386e-01 2.67480403e-01 6.62648261e-01
-1.61918104e-01 2.46356413e-01 -1.52324462e+00 4.08880621e-01
4.06021327e-01 -1.15014994e+00 1.57497466e-01 -1.61727443e-02
6.78439617e-01 1.51027247e-01 1.58224940e-01 7.38861859e-01
7.21204162e-01 -1.03423095e+00 8.74017775e-01 6.00375772e-01
1.03573728e+00 -4.06536460e-01 7.85755694e-01 1.52179524e-01
-8.78385901e-01 -2.29709037e-02 -5.10052741e-01 -1.24447584e-01
2.11048827e-01 3.59707355e-01 -7.91901052e-01 6.00687385e-01
4.45571423e-01 7.81354845e-01 -6.85530961e-01 9.42748487e-01
-4.36037213e-01 7.97772288e-01 1.54582977e-01 -3.69061440e-01
2.36869812e-01 2.41017848e-01 3.06635678e-01 1.48163712e+00
3.68952513e-01 -1.36401772e-01 -5.67111559e-02 3.01352203e-01
-1.84518918e-01 2.20825195e-01 -6.59024179e-01 6.62672892e-02
3.63981336e-01 9.76665020e-01 -2.75917321e-01 -8.03175569e-01
-3.42111021e-01 1.06018722e+00 7.19444752e-01 3.73441666e-01
-6.06456220e-01 -3.51864994e-01 -2.42853723e-02 -5.25333732e-02
4.17259216e-01 -7.24729076e-02 3.55028182e-01 -1.26148856e+00
7.15121180e-02 -8.36165607e-01 3.10937375e-01 -8.78843904e-01
-1.37018299e+00 7.14509130e-01 2.96666354e-01 -1.40536916e+00
-9.60692167e-01 -6.07308567e-01 -5.24028301e-01 1.04075789e+00
-1.82532346e+00 -9.36964214e-01 4.75094795e-01 2.66236007e-01
6.91016555e-01 -2.41671458e-01 1.07886994e+00 3.40819120e-01
1.74149334e-01 5.15902638e-01 7.63927042e-01 1.27161900e-02
9.85831618e-01 -1.38678181e+00 5.07818758e-01 7.48190820e-01
5.55649042e-01 8.06782544e-01 6.53514326e-01 -4.02470052e-01
-8.88757229e-01 -8.50594819e-01 1.64280796e+00 -1.02307904e+00
6.23762906e-01 -4.98552114e-01 -6.84382141e-01 7.19151914e-01
2.70944476e-01 -2.40974918e-01 8.31938088e-01 6.18518174e-01
-6.34053230e-01 1.45001799e-01 -4.02117044e-01 6.02772593e-01
1.12632251e+00 -9.52062964e-01 -1.06371403e+00 6.62313998e-01
6.74446881e-01 6.81829229e-02 -6.93053544e-01 3.53064567e-01
7.43384302e-01 -7.36737251e-01 1.17392504e+00 -1.04463446e+00
9.67170656e-01 -9.74591151e-02 -3.91241789e-01 -1.57149553e+00
-5.97157776e-01 -3.07776690e-01 -5.46687562e-03 1.10354245e+00
7.46063769e-01 -2.02345937e-01 4.59347039e-01 1.85476586e-01
-4.33754146e-01 -4.46158707e-01 -9.60666716e-01 -1.05932748e+00
4.02017713e-01 -3.47551733e-01 2.34486654e-01 6.62066519e-01
2.38793120e-01 9.08990204e-01 -4.77838606e-01 -3.42808276e-01
4.39674616e-01 -3.15174699e-01 3.46970081e-01 -1.24889636e+00
-6.39778972e-01 -3.80462110e-01 -2.03361779e-01 -1.22624040e+00
3.00476968e-01 -1.23602712e+00 4.70043063e-01 -1.63366783e+00
5.10146856e-01 -4.36007649e-01 -8.61844242e-01 3.19800049e-01
-3.80380243e-01 4.93169755e-01 5.35320155e-02 3.53430539e-01
-1.05539215e+00 5.22966266e-01 7.86661446e-01 -6.68453202e-02
2.17259780e-01 -1.95549726e-01 -8.24173093e-01 2.83076525e-01
5.62356174e-01 -7.03663170e-01 -5.61788499e-01 -4.24986601e-01
5.59302747e-01 1.65727139e-01 2.64052063e-01 -6.28769875e-01
1.08293511e-01 1.34158060e-01 2.16419041e-01 -3.74359131e-01
4.60058868e-01 -3.30383360e-01 -2.27093279e-01 3.58840883e-01
-1.26362133e+00 1.10404730e-01 -3.30577865e-02 6.10396862e-01
-3.16372693e-01 -7.67881930e-01 2.77340263e-01 -1.35270059e-01
-8.14587891e-01 -2.01694846e-01 -5.95234215e-01 4.73539196e-02
3.16948056e-01 1.60449356e-01 -5.97993791e-01 -7.18831480e-01
-4.45797890e-01 -6.67452663e-02 2.09201321e-01 6.34406388e-01
4.96265918e-01 -1.49569654e+00 -1.21746230e+00 -3.00870657e-01
6.83710337e-01 -7.80815721e-01 -1.09756783e-01 3.83737236e-01
6.60251155e-02 1.00585008e+00 1.61956146e-01 -2.84230053e-01
-1.17312014e+00 4.88671929e-01 3.03716375e-03 -9.03216183e-01
-2.12790847e-01 6.45362794e-01 1.97622284e-01 -5.03724098e-01
7.09068701e-02 1.93865314e-01 -3.95003408e-01 -4.92906244e-03
6.87558651e-01 -3.86150442e-02 2.35864729e-01 -4.51978534e-01
-1.98049232e-01 2.89336771e-01 -5.83647430e-01 -8.17768753e-01
1.11593413e+00 -2.96220668e-02 1.22434154e-01 7.46133447e-01
1.28008711e+00 -4.55621481e-02 -7.56243944e-01 -4.14996445e-01
3.52717102e-01 -4.65557098e-01 -3.19637209e-02 -1.25027907e+00
-1.73702106e-01 9.41732407e-01 4.04128373e-01 1.09819323e-01
8.11824739e-01 9.80440080e-02 3.02374452e-01 6.01952732e-01
5.06792665e-01 -9.97116923e-01 3.51448119e-01 7.51460850e-01
9.47957337e-01 -1.04774272e+00 4.18571662e-03 -1.93495452e-02
-5.94196558e-01 8.54358196e-01 1.89711064e-01 -8.80112052e-02
2.06577972e-01 -2.85340697e-01 -3.96634750e-02 -1.42560437e-01
-1.35285699e+00 -3.07355046e-01 6.78240657e-01 5.74491501e-01
1.02886176e+00 7.53264350e-04 -5.40166676e-01 2.92399287e-01
-2.63485551e-01 1.17844142e-01 4.82288271e-01 9.24688458e-01
-5.31572282e-01 -1.54081929e+00 -2.74406932e-02 6.14817083e-01
-6.28435135e-01 -8.33314002e-01 -1.95549428e-01 5.77590108e-01
-4.71596032e-01 9.73218083e-01 4.99707088e-02 -4.16997284e-01
1.45160824e-01 5.23065031e-01 5.33071101e-01 -1.07109690e+00
-6.14964783e-01 -8.29750076e-02 7.45792925e-01 -1.97915331e-01
-6.22230470e-01 -6.12997234e-01 -5.70399880e-01 1.93102416e-02
-5.22922695e-01 5.89832842e-01 6.36951864e-01 9.79890764e-01
1.02239259e-01 5.15363872e-01 7.27951229e-01 -7.26045430e-01
-1.10741770e+00 -1.31741703e+00 -4.11738306e-01 5.44598103e-01
1.96909741e-01 -5.07874846e-01 -2.90983915e-01 9.13092270e-02] | [11.46367073059082, 7.7584099769592285] |
503a8277-36b4-4b79-b615-e3a0b7543e7c | variational-relational-point-completion-1 | 2304.09131 | null | https://arxiv.org/abs/2304.09131v1 | https://arxiv.org/pdf/2304.09131v1.pdf | Variational Relational Point Completion Network for Robust 3D Classification | Real-scanned point clouds are often incomplete due to viewpoint, occlusion, and noise, which hampers 3D geometric modeling and perception. Existing point cloud completion methods tend to generate global shape skeletons and hence lack fine local details. Furthermore, they mostly learn a deterministic partial-to-complete mapping, but overlook structural relations in man-made objects. To tackle these challenges, this paper proposes a variational framework, Variational Relational point Completion Network (VRCNet) with two appealing properties: 1) Probabilistic Modeling. In particular, we propose a dual-path architecture to enable principled probabilistic modeling across partial and complete clouds. One path consumes complete point clouds for reconstruction by learning a point VAE. The other path generates complete shapes for partial point clouds, whose embedded distribution is guided by distribution obtained from the reconstruction path during training. 2) Relational Enhancement. Specifically, we carefully design point self-attention kernel and point selective kernel module to exploit relational point features, which refines local shape details conditioned on the coarse completion. In addition, we contribute multi-view partial point cloud datasets (MVP and MVP-40 dataset) containing over 200,000 high-quality scans, which render partial 3D shapes from 26 uniformly distributed camera poses for each 3D CAD model. Extensive experiments demonstrate that VRCNet outperforms state-of-the-art methods on all standard point cloud completion benchmarks. Notably, VRCNet shows great generalizability and robustness on real-world point cloud scans. Moreover, we can achieve robust 3D classification for partial point clouds with the help of VRCNet, which can highly increase classification accuracy. | ['Ziwei Liu', 'Shuai Yi', 'Haiyu Zhao', 'Junzhe Zhang', 'Zhongang Cai', 'Xinyi Chen', 'Liang Pan'] | 2023-04-18 | null | null | null | null | ['point-cloud-completion', '3d-classification'] | ['computer-vision', 'computer-vision'] | [-2.02265695e-01 3.75304297e-02 -1.08945638e-01 -1.99978799e-01
-1.09926462e+00 -5.76152921e-01 6.63685262e-01 -2.56467730e-01
1.85023800e-01 1.14920035e-01 -5.61737567e-02 5.46856448e-02
-1.37257978e-01 -9.75106478e-01 -1.21672618e+00 -4.90961254e-01
4.09016758e-01 1.11599982e+00 1.95746168e-01 -4.28710915e-02
2.30775371e-01 9.53530312e-01 -1.38602591e+00 -1.94622912e-02
9.73958254e-01 8.05377007e-01 5.40073335e-01 2.26401359e-01
-5.45196056e-01 3.46961878e-02 -2.39658859e-02 -4.94732559e-01
3.18821043e-01 5.76770365e-01 -3.72997254e-01 3.07616413e-01
6.76097155e-01 -3.94808501e-01 -5.49432337e-01 9.30404365e-01
1.77824080e-01 3.90707282e-03 9.68200326e-01 -1.19440687e+00
-8.85399401e-01 9.99527648e-02 -9.09503579e-01 -5.63251436e-01
2.58631945e-01 2.53880143e-01 9.87485766e-01 -1.59263742e+00
6.84549451e-01 1.51193976e+00 7.55295753e-01 4.78895783e-01
-1.46858156e+00 -7.54320443e-01 2.70309865e-01 -3.33846360e-01
-1.67721486e+00 -4.55655865e-02 1.19252563e+00 -4.81762767e-01
7.77319729e-01 9.34354588e-02 6.68201745e-01 1.11320055e+00
1.48294289e-02 7.91821003e-01 6.92779064e-01 1.42279744e-01
1.37740150e-01 -1.91939518e-01 -3.14624965e-01 4.76727784e-01
1.57963023e-01 1.88418880e-01 -4.40945923e-01 -5.38741946e-01
1.36581588e+00 5.20819485e-01 -3.51694047e-01 -1.00815618e+00
-1.46874428e+00 6.13035142e-01 5.47358453e-01 -1.93029478e-01
-6.20094299e-01 2.99547136e-01 -7.51003399e-02 -1.31838322e-01
5.36627829e-01 -4.58170623e-02 -4.26372170e-01 1.55461624e-01
-6.79300189e-01 4.75269049e-01 4.01284665e-01 1.66647017e+00
1.08398080e+00 -8.24740902e-03 1.66068494e-01 8.38198245e-01
6.59919322e-01 1.07553971e+00 -3.41008991e-01 -1.27579629e+00
7.61833072e-01 7.08880007e-01 1.29354000e-01 -1.00451756e+00
2.34853085e-02 -3.80793929e-01 -9.89801168e-01 2.12313071e-01
-5.84436990e-02 5.38712442e-01 -9.40906823e-01 1.36409366e+00
6.54292822e-01 5.39861381e-01 -1.74014047e-01 8.58043492e-01
9.35567081e-01 6.82039320e-01 -1.61627844e-01 1.96065158e-01
1.13022220e+00 -5.49971521e-01 -1.27024278e-01 7.39611313e-02
2.91637145e-02 -6.81109726e-01 1.08233213e+00 3.10462624e-01
-1.15753579e+00 -5.77215374e-01 -8.02222490e-01 -3.86391997e-01
2.72531718e-01 -8.56545866e-02 5.64505577e-01 1.49340019e-01
-7.76900709e-01 6.48355842e-01 -1.16708159e+00 1.27622947e-01
8.69621098e-01 2.16537967e-01 -4.85390186e-01 -6.36328876e-01
-2.77529806e-01 1.64263383e-01 -1.74124286e-01 -6.97311983e-02
-1.08161175e+00 -1.39364541e+00 -8.46679509e-01 4.12217118e-02
2.03786388e-01 -1.26316869e+00 9.68651950e-01 -7.86074549e-02
-1.27562273e+00 9.32819664e-01 -3.37533951e-01 1.93367586e-01
6.84305727e-01 -2.00635538e-01 3.37972306e-02 1.46901131e-01
3.52038294e-01 8.88189077e-01 1.02602279e+00 -1.88942122e+00
-3.30742776e-01 -6.56215072e-01 -2.35095143e-01 2.69314826e-01
2.98550755e-01 -6.41692638e-01 -9.46365893e-01 -6.07032895e-01
8.63350630e-01 -8.21133852e-01 -2.39721671e-01 3.20894629e-01
-4.45593774e-01 -4.41492409e-01 8.31633389e-01 -2.61240840e-01
2.31488854e-01 -2.23521900e+00 3.01191241e-01 1.64319694e-01
4.29152519e-01 -3.58485818e-01 -2.28871748e-01 2.62228847e-01
-2.51095872e-02 1.98250800e-01 -3.73813391e-01 -1.00422370e+00
2.81285256e-01 5.50942600e-01 -7.89607704e-01 5.20573437e-01
3.24781120e-01 1.05656970e+00 -7.55891085e-01 -4.35637087e-01
6.02326691e-01 9.24510300e-01 -6.26795530e-01 1.62373602e-01
-5.28544009e-01 5.86189628e-01 -8.07723343e-01 9.93825018e-01
1.33921325e+00 -3.57585996e-01 -5.41508913e-01 -4.17817712e-01
4.55366336e-02 -4.10137698e-02 -1.10518324e+00 2.29846263e+00
-4.26804006e-01 3.13393250e-02 2.05282167e-01 -4.22304571e-01
1.03671563e+00 3.09522986e-01 8.17468345e-01 -1.43058196e-01
-1.51268408e-01 1.68176368e-01 -7.98071146e-01 6.70980848e-03
4.63458657e-01 -2.10084841e-01 3.25135291e-01 1.52026877e-01
3.85439023e-02 -9.00012136e-01 -6.54154241e-01 3.53482753e-01
8.51317227e-01 6.09911084e-01 -3.26700389e-01 8.13337788e-02
2.74891585e-01 4.67096940e-02 6.84217989e-01 3.56674165e-01
2.55588621e-01 1.36263835e+00 2.17381597e-01 -3.74185145e-01
-1.22409582e+00 -1.73245072e+00 -3.26758116e-01 2.48996645e-01
5.02922237e-01 -2.70711154e-01 -3.09434533e-01 -4.69684094e-01
3.12567770e-01 6.62008166e-01 -3.15828413e-01 4.46025729e-02
-5.94174922e-01 -3.67752612e-01 9.47532654e-02 5.47230244e-01
1.85987309e-01 -8.62482369e-01 2.49541961e-02 8.45316872e-02
-2.76244253e-01 -1.32301128e+00 -4.86409903e-01 -4.43724483e-01
-1.40919077e+00 -1.03632915e+00 -7.22720921e-01 -4.84955311e-01
7.48356581e-01 7.75047779e-01 1.43495023e+00 2.16029100e-02
6.29277751e-02 8.66277754e-01 -1.89033419e-01 -3.06776792e-01
-1.11289233e-01 -4.06652465e-02 5.63041717e-02 6.19102269e-02
2.38477334e-01 -1.20798123e+00 -6.18700981e-01 5.12267113e-01
-7.85176039e-01 1.41760051e-01 7.18680382e-01 6.34412169e-01
1.50398922e+00 -9.67483893e-02 -3.22868191e-02 -5.52589417e-01
-1.03441365e-01 -5.87021589e-01 -7.07983315e-01 9.80226696e-03
-2.38848761e-01 -1.89348534e-01 1.53404474e-01 -2.92196602e-01
-9.59626436e-01 2.90771067e-01 -2.70068437e-01 -1.54027236e+00
-1.98328093e-01 4.36058976e-02 -5.20224154e-01 -2.16004744e-01
4.05931145e-01 5.31192482e-01 -4.47105207e-02 -7.86844969e-01
3.90673101e-01 7.40991011e-02 6.39167130e-01 -1.01556194e+00
1.41021824e+00 1.01948392e+00 2.19726384e-01 -9.06820416e-01
-3.66595417e-01 -4.71270114e-01 -7.90041685e-01 -6.41853036e-03
7.74759233e-01 -1.34323049e+00 -7.19708085e-01 2.19553158e-01
-1.47560501e+00 -8.78009200e-02 -3.05942565e-01 4.57204312e-01
-8.13269854e-01 5.57758212e-01 -4.22220826e-01 -6.68313384e-01
-2.15982899e-01 -1.17843008e+00 1.82218146e+00 -1.65863916e-01
3.26269418e-01 -6.47205055e-01 -6.09674156e-02 4.15043384e-01
-2.15805143e-01 3.76408964e-01 8.18912387e-01 1.33304909e-01
-1.54119611e+00 -1.01319522e-01 -3.00719708e-01 2.72153407e-01
-3.71961761e-03 1.41360044e-01 -1.02559304e+00 -1.80590823e-01
1.66670829e-01 -1.35475501e-01 7.17627764e-01 5.04886925e-01
1.30361581e+00 1.58366650e-01 -5.00750780e-01 1.09765768e+00
1.45479715e+00 -5.16651452e-01 6.57937109e-01 -2.51561463e-01
1.33246911e+00 5.27611852e-01 6.37164414e-01 4.48708326e-01
6.86473191e-01 6.27325356e-01 1.07327843e+00 1.12751625e-01
-9.11271200e-02 -8.31883669e-01 6.99033262e-03 8.78985047e-01
-3.79089862e-01 1.11541212e-01 -8.90050769e-01 4.84364152e-01
-1.59824872e+00 -6.40978575e-01 -5.17306149e-01 2.18251824e+00
3.48815113e-01 -2.59798821e-02 -3.10194284e-01 -2.50201344e-01
5.08233249e-01 2.16670319e-01 -6.88962340e-01 6.40109479e-01
-1.06376044e-01 1.96598977e-01 3.19646358e-01 4.08963352e-01
-7.31117725e-01 9.39330518e-01 4.70396137e+00 1.08389938e+00
-6.99265182e-01 2.23581567e-01 2.71211982e-01 4.50901911e-02
-9.88417208e-01 8.23296458e-02 -6.69770777e-01 2.66276419e-01
3.12258173e-02 2.46403784e-01 3.35061222e-01 1.18557131e+00
-9.63771418e-02 4.88568753e-01 -1.11261213e+00 1.51220071e+00
-2.37452626e-01 -1.59750009e+00 4.32026505e-01 4.38547820e-01
8.64785254e-01 6.05304539e-01 1.32945225e-01 1.71356201e-01
3.70739996e-01 -9.14830446e-01 9.21542466e-01 7.86684155e-01
9.15348887e-01 -8.46522808e-01 2.81683475e-01 5.93395710e-01
-1.39629471e+00 4.78108376e-01 -8.84199560e-01 4.32269305e-01
3.76236767e-01 7.94439375e-01 -3.75359803e-01 1.06794667e+00
7.61393964e-01 9.75494027e-01 -2.03522339e-01 9.30839360e-01
-1.24189563e-01 2.69788355e-01 -5.81521928e-01 5.20854712e-01
8.29081386e-02 -5.44206023e-01 9.23210800e-01 4.81735498e-01
5.77464521e-01 2.81268150e-01 2.01001465e-01 1.56962287e+00
-1.43399760e-01 -1.80995181e-01 -6.73549831e-01 3.63735437e-01
7.25939035e-01 1.27020359e+00 -5.13601482e-01 -5.20897703e-03
-3.91387433e-01 7.83494651e-01 3.76016259e-01 4.94051129e-01
-6.38400495e-01 2.98673481e-01 9.31432605e-01 2.61543244e-01
3.96171391e-01 -5.69628060e-01 -5.87574363e-01 -1.33003175e+00
3.63115788e-01 -2.82891840e-01 -2.02713683e-01 -1.09549093e+00
-1.65604341e+00 4.42442834e-01 1.13782018e-01 -1.59345710e+00
2.88333803e-01 -5.09476840e-01 -5.30908167e-01 1.02359343e+00
-1.48048532e+00 -1.59718263e+00 -6.04339302e-01 6.95933580e-01
5.96287072e-01 5.12663461e-02 4.18941826e-01 1.88045278e-01
-6.63713273e-03 1.14867195e-01 -2.11799219e-01 -1.41065806e-01
3.53714257e-01 -1.13584840e+00 8.28039050e-01 4.11370665e-01
4.42883700e-01 5.73461533e-01 2.77399242e-01 -8.35374057e-01
-1.88042724e+00 -1.33464277e+00 2.90712267e-01 -1.09790027e+00
1.16050743e-01 -4.42791998e-01 -1.22081721e+00 7.05883026e-01
-5.08851945e-01 3.69373500e-01 1.00374095e-01 -8.35180357e-02
-6.05279803e-01 6.46089688e-02 -1.05059052e+00 5.50777793e-01
1.37931299e+00 -5.70757449e-01 -5.95409036e-01 3.92430156e-01
1.15995860e+00 -7.82708347e-01 -1.00469077e+00 7.04992235e-01
2.26342812e-01 -9.13365364e-01 1.52133334e+00 -1.86465442e-01
6.76204383e-01 -4.99858677e-01 -5.57872891e-01 -1.06870389e+00
-4.14803207e-01 -5.41608512e-01 -2.59088159e-01 1.24681914e+00
1.11643923e-02 -4.56411034e-01 1.10474598e+00 5.17456710e-01
-7.23373353e-01 -8.98943841e-01 -9.90745544e-01 -7.35680103e-01
3.01769674e-01 -9.93570924e-01 1.16664231e+00 9.50159371e-01
-9.64393973e-01 1.52563125e-01 -9.80654582e-02 6.43171012e-01
1.17259681e+00 3.67445797e-01 1.30138505e+00 -1.43600357e+00
-1.57243699e-01 -4.14812922e-01 -3.90809029e-01 -1.55687141e+00
1.95286885e-01 -8.94724965e-01 -2.03005672e-01 -1.48333597e+00
6.98270323e-03 -9.69063759e-01 2.13130876e-01 9.38710049e-02
-4.97674942e-02 2.17691049e-01 8.47952738e-02 6.48205876e-01
-1.41966283e-01 1.19991624e+00 1.54806161e+00 -1.59825251e-01
-1.68543190e-01 1.89067379e-01 -4.22279507e-01 7.80333459e-01
3.03527117e-01 -5.53608775e-01 -4.79826629e-01 -8.16091418e-01
2.03279123e-01 2.64587402e-01 9.00920391e-01 -8.35162640e-01
7.83102885e-02 -3.49057794e-01 4.59634632e-01 -1.43479824e+00
9.11484122e-01 -1.08706582e+00 5.53533316e-01 -1.28054172e-01
3.70478660e-01 -1.01427883e-01 -6.50319159e-02 1.09812081e+00
-3.33825164e-02 1.53588951e-01 5.57134688e-01 -2.46637344e-01
-2.75388688e-01 1.24629879e+00 6.62974119e-01 -5.50354570e-02
7.42838621e-01 -3.14228743e-01 1.62267730e-01 -3.01254261e-02
-5.13871729e-01 3.56941074e-01 1.02532768e+00 5.11455178e-01
1.12485611e+00 -1.66012979e+00 -8.01033556e-01 3.92948002e-01
3.89059156e-01 1.22007084e+00 7.42083669e-01 6.32948041e-01
-6.28816545e-01 1.30403936e-01 2.06749216e-01 -1.43479955e+00
-7.19426394e-01 6.26382053e-01 1.45316586e-01 2.50745296e-01
-1.31531882e+00 8.15294147e-01 7.11317956e-01 -1.02087176e+00
-3.45163010e-02 -5.42657435e-01 3.42821002e-01 -5.14051318e-01
2.04417720e-01 2.49984697e-01 1.73567440e-02 -7.47057557e-01
-1.44634083e-01 1.13484228e+00 8.89867395e-02 -1.01903893e-01
1.46488929e+00 -5.49147697e-03 -6.50460646e-02 2.75115341e-01
1.04602075e+00 1.94859728e-01 -1.73283529e+00 -4.84929472e-01
-4.84056711e-01 -9.34634030e-01 5.01068756e-02 -1.59776047e-01
-1.13290429e+00 1.01648533e+00 1.18542984e-01 -3.18200320e-01
5.40427446e-01 3.92008483e-01 7.54630864e-01 1.99309394e-01
9.22582328e-01 -4.37463611e-01 -1.10445749e-02 3.96680892e-01
1.20551991e+00 -1.22052205e+00 1.17534757e-01 -8.74424279e-01
-6.10341668e-01 8.55851412e-01 5.48759758e-01 -4.92659628e-01
7.93987393e-01 -1.86464623e-01 -3.25160295e-01 -6.36690199e-01
-6.35261893e-01 3.02353382e-01 4.13602322e-01 9.55094337e-01
-3.39006364e-01 1.66040495e-01 5.67033648e-01 5.79237223e-01
-2.90468007e-01 -2.17590943e-01 9.56601724e-02 4.98761058e-01
-9.49936286e-02 -9.14033771e-01 -6.62856042e-01 2.93218195e-01
1.75961956e-01 1.40268043e-01 -1.32451177e-01 8.05076599e-01
-7.32335299e-02 3.45153481e-01 2.63507336e-01 -3.63784820e-01
5.19181788e-01 -2.76102334e-01 5.29928327e-01 -6.19129598e-01
1.29652053e-01 3.03800434e-01 -5.94259858e-01 -8.23133409e-01
-1.48726046e-01 -8.89563799e-01 -1.27759981e+00 -4.22866553e-01
-6.18192106e-02 -7.31317997e-02 6.56964183e-01 6.16991997e-01
7.00760305e-01 3.34991693e-01 6.25587165e-01 -1.47991753e+00
-5.74161708e-01 -6.93367302e-01 -6.68590963e-01 3.81343037e-01
2.30799466e-01 -1.03093815e+00 -3.08689386e-01 -2.03324527e-01] | [8.39339542388916, -3.5358898639678955] |
4334556a-ebe3-40f9-b1fe-a75af9ce6493 | neural-text-generation-from-structured-data | 1603.07771 | null | http://arxiv.org/abs/1603.07771v3 | http://arxiv.org/pdf/1603.07771v3.pdf | Neural Text Generation from Structured Data with Application to the Biography Domain | This paper introduces a neural model for concept-to-text generation that
scales to large, rich domains. We experiment with a new dataset of biographies
from Wikipedia that is an order of magnitude larger than existing resources
with over 700k samples. The dataset is also vastly more diverse with a 400k
vocabulary, compared to a few hundred words for Weathergov or Robocup. Our
model builds upon recent work on conditional neural language model for text
generation. To deal with the large vocabulary, we extend these models to mix a
fixed vocabulary with copy actions that transfer sample-specific words from the
input database to the generated output sentence. Our neural model significantly
out-performs a classical Kneser-Ney language model adapted to this task by
nearly 15 BLEU. | ['Michael Auli', 'David Grangier', 'Remi Lebret'] | 2016-03-24 | neural-text-generation-from-structured-data-1 | https://aclanthology.org/D16-1128 | https://aclanthology.org/D16-1128.pdf | emnlp-2016-11 | ['table-to-text-generation', 'concept-to-text-generation'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.30649397e-01 5.48144162e-01 -2.44130626e-01 -3.07119697e-01
-1.20793462e+00 -6.15632534e-01 1.25549972e+00 -2.70976359e-03
-4.60381448e-01 1.47952175e+00 8.06713402e-01 -2.91867673e-01
2.68471658e-01 -1.29606128e+00 -9.27316606e-01 -2.04375833e-01
1.61803484e-01 1.06222105e+00 -7.72841796e-02 -7.63440609e-01
3.05293560e-01 -1.25116156e-02 -1.28459501e+00 5.21987200e-01
8.29118311e-01 3.34736764e-01 1.51435390e-01 9.99916732e-01
-6.00143254e-01 8.25402856e-01 -1.00404823e+00 -5.57498395e-01
9.09011904e-03 -3.60838503e-01 -1.16028404e+00 -4.74673122e-01
7.77318656e-01 -4.65143323e-01 -4.20158327e-01 5.99740028e-01
7.31058538e-01 2.36547947e-01 1.08415437e+00 -8.47243965e-01
-1.13085341e+00 1.35429335e+00 -4.79559861e-02 1.70974851e-01
4.22170132e-01 3.12136230e-03 1.04214227e+00 -8.18296552e-01
9.63210046e-01 1.39631116e+00 4.52300340e-01 1.06633282e+00
-1.09653211e+00 -7.39845157e-01 -3.15887667e-02 -3.48703712e-01
-1.26866972e+00 -4.90787476e-01 3.53760183e-01 -3.74222726e-01
1.51987278e+00 1.42578155e-01 5.36547244e-01 1.59627104e+00
2.41914123e-01 7.33925819e-01 5.71924448e-01 -5.90254724e-01
1.96512237e-01 1.32292002e-01 9.68724191e-02 3.20865184e-01
3.49497139e-01 -2.28241216e-02 -7.41115391e-01 -2.97154397e-01
6.55222356e-01 -4.88629639e-01 1.47032216e-02 5.33210002e-02
-1.41221857e+00 1.12669802e+00 1.80981979e-01 1.96886837e-01
-1.84795797e-01 4.83557314e-01 3.42777491e-01 3.29980999e-01
6.76974654e-01 8.46480429e-01 -4.57835346e-01 -1.78526774e-01
-1.12165821e+00 9.01488841e-01 1.29585671e+00 1.35791016e+00
5.07436097e-01 2.88424820e-01 -6.13487244e-01 8.07402849e-01
1.32259622e-01 6.41175508e-01 9.35637236e-01 -6.08781278e-01
8.39352190e-01 6.32925630e-02 2.33655185e-01 -5.46338677e-01
-4.81333695e-02 -1.40289813e-01 -9.12538469e-01 -2.24093527e-01
3.08367968e-01 -3.84623289e-01 -1.06891894e+00 1.58807039e+00
-3.36577408e-02 -1.35517687e-01 3.02162468e-01 3.70702714e-01
1.06743443e+00 1.13176191e+00 1.57306552e-01 1.64200917e-01
1.20639467e+00 -9.56003904e-01 -4.90840197e-01 -5.93172371e-01
5.56830108e-01 -5.93551993e-01 1.01603770e+00 1.99358448e-01
-1.16525185e+00 -4.28307027e-01 -1.03376484e+00 -3.74811620e-01
-7.61322737e-01 -1.59420781e-02 7.45707035e-01 5.92794776e-01
-1.32908654e+00 4.44111049e-01 -2.54588574e-01 -6.66470349e-01
2.04976156e-01 -1.32232131e-02 -1.56673402e-01 2.60290448e-02
-1.69075835e+00 9.30460215e-01 1.02800620e+00 -5.50077200e-01
-1.15150166e+00 -7.46449649e-01 -8.72423410e-01 -6.51857033e-02
1.49876669e-01 -1.01267183e+00 1.52279234e+00 -6.16118670e-01
-1.68052924e+00 6.27024531e-01 -1.03222251e-01 -7.80437469e-01
3.11474532e-01 -1.48060068e-01 -3.16728443e-01 -1.44072324e-01
1.21028900e-01 1.04769993e+00 6.64952219e-01 -9.65369463e-01
-3.00894022e-01 1.37552500e-01 1.33988023e-01 3.66713256e-01
-4.64061916e-01 -1.32941663e-01 -2.82265216e-01 -1.11651170e+00
-9.07741010e-01 -5.30313134e-01 -3.63149762e-01 -4.65250999e-01
-5.06844997e-01 -3.72459471e-01 7.30330572e-02 -7.01469302e-01
1.24619675e+00 -1.43907249e+00 1.07212745e-01 -1.28086537e-01
1.30136669e-01 7.98154250e-02 -5.91854215e-01 8.82618785e-01
1.42238513e-01 3.42263609e-01 -1.53338999e-01 -6.16030842e-02
8.46661329e-02 -2.14866139e-02 -7.07677364e-01 -3.12074035e-01
3.17814797e-01 1.20380342e+00 -1.10956573e+00 -4.11718339e-01
-2.67985642e-01 1.41333297e-01 -6.53255999e-01 7.89928660e-02
-7.12210953e-01 -1.96007937e-01 -5.11840999e-01 1.59122780e-01
2.40108684e-01 -3.44838440e-01 4.65146676e-02 4.57409859e-01
1.75685421e-01 7.63214707e-01 -7.70537138e-01 1.89708722e+00
-6.81108952e-01 6.94154978e-01 -4.10386026e-01 -7.02316880e-01
1.06952274e+00 4.44946378e-01 -1.60404250e-01 -2.01720893e-01
-8.20720270e-02 6.10220321e-02 -3.22565645e-01 -1.13956153e-01
1.24777496e+00 -3.48144174e-01 -3.69990051e-01 6.93834484e-01
4.30085421e-01 -7.05061674e-01 6.87075615e-01 7.95201898e-01
9.67201114e-01 5.89008071e-02 4.49682385e-01 -2.22778022e-01
1.78454876e-01 6.08424209e-02 -3.55630554e-02 1.09943283e+00
5.10576189e-01 4.90150690e-01 3.75757754e-01 -4.27352697e-01
-1.32186294e+00 -1.21444857e+00 2.44458690e-01 1.20921409e+00
-3.79683882e-01 -6.07644439e-01 -8.06384206e-01 -6.00703061e-01
1.38370804e-02 1.27431524e+00 -5.90199411e-01 -1.17669471e-01
-4.66768920e-01 -7.96957850e-01 1.06410658e+00 6.23201370e-01
2.87338674e-01 -1.53776348e+00 6.78671524e-02 4.09718424e-01
-4.47497368e-01 -8.81696939e-01 -5.02397597e-01 -8.73888843e-03
-7.28433549e-01 -4.50654060e-01 -1.04730475e+00 -9.03517008e-01
5.22553205e-01 -3.46753508e-01 1.85423803e+00 -3.83523673e-01
-3.77721846e-01 1.47206515e-01 -3.88712376e-01 -7.51672447e-01
-1.08975124e+00 6.00751936e-01 -8.12489018e-02 -6.41467750e-01
5.03531694e-01 -1.60252646e-01 5.08081838e-02 -5.13639212e-01
-1.17270648e+00 2.57199496e-01 5.95225036e-01 9.51837838e-01
1.27496734e-01 -3.61139446e-01 9.95932698e-01 -9.21479285e-01
1.23522186e+00 -7.20377505e-01 -4.01616305e-01 3.26427221e-01
-4.69054878e-01 3.75950783e-01 6.97829187e-01 -5.12540281e-01
-1.16084754e+00 -1.18158467e-01 -8.55404213e-02 2.74194330e-01
-4.00891490e-02 5.81628025e-01 1.48698911e-01 5.37812173e-01
1.01547873e+00 4.60296392e-01 -1.86878309e-01 -2.52377868e-01
9.37329888e-01 9.30128515e-01 3.74948233e-01 -6.15906537e-01
8.45395565e-01 5.39180078e-02 -5.55620670e-01 -8.57628405e-01
-6.53578937e-01 -8.25994834e-02 -4.61188227e-01 3.16248462e-02
8.10006022e-01 -1.18696964e+00 -1.68253273e-01 5.34898102e-01
-1.39594710e+00 -7.28398502e-01 -5.33826828e-01 2.73472399e-01
-5.62228858e-01 2.78053492e-01 -8.47222090e-01 -5.94737053e-01
-8.98860335e-01 -4.92350519e-01 1.11051381e+00 1.51652992e-01
-4.63224202e-01 -1.14527071e+00 3.54875147e-01 2.03042269e-01
6.05495751e-01 -6.68778792e-02 9.43692982e-01 -7.93606937e-01
-4.14975405e-01 -2.29845792e-01 -4.46503870e-02 2.76768893e-01
-1.25959948e-01 -3.18874866e-01 -6.67819619e-01 -1.40129879e-01
-6.12449408e-01 -9.18952823e-01 1.19822943e+00 1.12736367e-01
9.23275054e-01 -4.77183193e-01 -3.54452968e-01 1.40166506e-01
1.23781180e+00 -1.46754876e-01 6.74465179e-01 1.58189639e-01
5.50116837e-01 5.12832463e-01 2.90462136e-01 3.54064018e-01
5.05548120e-01 3.17383170e-01 -4.90557821e-03 1.66152865e-01
-1.55789152e-01 -6.71089232e-01 6.09183252e-01 9.78861928e-01
1.27017483e-01 -8.75378311e-01 -9.39739585e-01 9.76545572e-01
-1.44282341e+00 -1.17282104e+00 1.70247301e-01 1.98580897e+00
1.52631712e+00 1.61584422e-01 -1.34763494e-01 -5.08663833e-01
5.18873632e-01 3.94602448e-01 -3.35382342e-01 -6.21641338e-01
-1.31010801e-01 5.20700157e-01 6.77887261e-01 7.47811913e-01
-8.97205770e-01 1.35349047e+00 7.68417025e+00 1.19373333e+00
-7.02584326e-01 -2.10462213e-01 5.10437787e-01 -4.77575094e-01
-6.79291785e-01 -1.71730995e-01 -1.22638154e+00 3.67782235e-01
1.47225702e+00 -7.81585395e-01 6.73625112e-01 8.14521849e-01
-1.55584157e-01 2.28151947e-01 -1.18274999e+00 6.88021779e-01
5.85411608e-01 -1.82000530e+00 8.30513656e-01 2.27347668e-02
9.86330271e-01 2.17460245e-01 -2.48914659e-01 7.00065911e-01
1.28334498e+00 -1.47846901e+00 6.36334598e-01 6.67177141e-01
1.12635207e+00 -6.75042927e-01 5.01670063e-01 4.88942891e-01
-6.90176606e-01 3.02850395e-01 -7.15022922e-01 -1.52690932e-01
3.99312228e-01 5.96809804e-01 -1.33948815e+00 3.34469348e-01
3.63626450e-01 6.03729665e-01 -5.91111362e-01 5.04078209e-01
-2.61167258e-01 6.28952384e-01 -1.57641619e-01 -6.68368936e-01
2.66063720e-01 2.70347089e-01 4.62501764e-01 1.42483914e+00
5.30083120e-01 -2.15197906e-01 2.37250812e-02 9.37348187e-01
-6.52989984e-01 3.08152854e-01 -1.03559613e+00 -4.74842489e-01
6.18114173e-01 1.11384571e+00 -3.54721218e-01 -9.48717058e-01
-5.08839712e-02 1.21172071e+00 5.39279878e-01 3.68251473e-01
-4.94006842e-01 -8.33314240e-01 3.46483141e-01 -1.64742954e-02
3.85938048e-01 -1.21006750e-01 -8.53771269e-02 -1.26256168e+00
-1.72716916e-01 -9.74464834e-01 1.39745250e-01 -9.49600339e-01
-1.63910675e+00 6.88321471e-01 1.55034691e-01 -7.53656864e-01
-9.62302208e-01 -5.11160851e-01 -5.83890259e-01 1.25651550e+00
-1.25992811e+00 -1.17683101e+00 -6.53827488e-02 2.93203980e-01
8.72594714e-01 -5.44218302e-01 1.21767509e+00 -1.08972818e-01
5.16182138e-03 5.33251166e-01 4.29375976e-01 4.49618042e-01
9.63649809e-01 -1.48942387e+00 1.43730104e+00 5.93850195e-01
9.75498557e-02 8.68451238e-01 6.00412786e-01 -9.04815495e-01
-9.91115272e-01 -1.36000884e+00 1.42911136e+00 -9.29763079e-01
6.91523969e-01 -7.86822796e-01 -4.24357414e-01 8.12302947e-01
6.88834310e-01 -6.60840750e-01 6.92036986e-01 8.49873424e-02
-4.74390626e-01 2.13531613e-01 -9.05912340e-01 7.82948971e-01
8.51795614e-01 -4.24383998e-01 -8.43599796e-01 5.19230902e-01
1.19473708e+00 -3.65977466e-01 -7.77561963e-01 -9.78609249e-02
5.10828793e-01 -2.35575497e-01 8.89561296e-01 -9.03271079e-01
9.69912708e-01 4.12329659e-02 5.56157157e-02 -1.70081055e+00
-4.47130352e-01 -5.67944407e-01 -1.77273512e-01 1.37646186e+00
8.30391943e-01 -3.05590481e-01 7.38503754e-01 3.90927315e-01
1.08590379e-01 -2.63094008e-01 -7.50065863e-01 -7.27772951e-01
8.88798654e-01 -3.17188144e-01 8.04855466e-01 7.12219059e-01
1.47998214e-01 9.47878540e-01 -6.18254185e-01 -5.43378949e-01
3.16476107e-01 -2.43985400e-01 8.78250003e-01 -1.22088444e+00
-4.20738846e-01 -3.73972386e-01 5.16529866e-02 -1.23433208e+00
2.70829856e-01 -1.35194552e+00 3.04813027e-01 -2.01205897e+00
2.59383529e-01 -1.77214712e-01 2.46147603e-01 4.52081203e-01
-3.09449583e-01 3.25988919e-01 1.00748554e-01 -2.88694412e-01
-3.56680691e-01 5.80270708e-01 1.03775263e+00 -4.77770209e-01
-4.78492789e-02 -4.31814611e-01 -1.14338481e+00 3.05667251e-01
8.41353178e-01 -3.16924840e-01 -5.46435475e-01 -7.57665992e-01
5.59467673e-01 -8.51479266e-03 6.39919862e-02 -7.78204620e-01
-8.47718678e-03 -1.80155978e-01 5.19061863e-01 -4.62438285e-01
2.10788310e-01 3.00863758e-02 -2.35492095e-01 1.85185388e-01
-1.10540462e+00 -1.08945072e-01 2.28403300e-01 6.96252167e-01
3.52991484e-02 -3.79998922e-01 3.62642825e-01 -6.08447373e-01
-5.49383938e-01 1.32905424e-01 -9.44549143e-01 5.93709588e-01
7.37166822e-01 1.09516442e-01 -5.15486836e-01 -8.43381703e-01
-3.82021785e-01 2.34863818e-01 2.58865267e-01 8.53970647e-01
5.10375261e-01 -1.33756518e+00 -1.25494480e+00 -1.03668971e-02
2.57511705e-01 1.85159538e-02 -2.64877044e-02 -1.75235212e-01
-6.02344811e-01 7.96497524e-01 -8.05609766e-03 1.55803159e-01
-8.27462077e-01 4.98619288e-01 2.72622425e-02 -7.43673980e-01
-4.37201023e-01 9.43762779e-01 -4.99901101e-02 -8.28541219e-01
9.85547900e-02 -3.68181944e-01 -3.10144156e-01 4.82625403e-02
9.11439002e-01 2.67979085e-01 -1.08355194e-01 -2.71121740e-01
5.02429977e-02 1.90644525e-02 -4.30481732e-01 -5.61003566e-01
1.09347594e+00 1.61991298e-01 -1.17597319e-01 5.74911594e-01
9.44957733e-01 1.01451278e-01 -5.80724299e-01 -1.64556369e-01
-1.59791589e-01 1.95611700e-01 -2.36528948e-01 -1.04863894e+00
-4.71538424e-01 7.70556867e-01 -8.44681263e-02 -2.67340206e-02
5.62765360e-01 1.09864905e-01 1.03280354e+00 8.21117103e-01
4.59248781e-01 -1.23381746e+00 1.82426304e-01 1.13537490e+00
1.03126061e+00 -8.29077661e-01 -5.28400652e-02 1.51892915e-01
-7.79397368e-01 1.03076839e+00 6.31324887e-01 -2.28236020e-01
4.63471532e-01 3.60391229e-01 -1.47094252e-02 2.71739904e-02
-1.16298676e+00 1.28386006e-01 2.22355112e-01 8.80129695e-01
7.16700375e-01 1.79488406e-01 -3.87564451e-01 5.59802830e-01
-8.66811454e-01 2.55729914e-01 9.23822820e-01 5.98838687e-01
-5.53089380e-01 -1.25948036e+00 -1.44991592e-01 8.26491714e-01
-4.12762940e-01 -8.66814077e-01 -6.73658133e-01 5.24966657e-01
-3.29424500e-01 9.41193640e-01 3.12857598e-01 -1.65640831e-01
-1.22764021e-01 4.97093886e-01 1.97732285e-01 -1.02198780e+00
-6.27413273e-01 -3.85273188e-01 4.89073426e-01 -2.52778023e-01
-9.97199416e-02 -3.51434469e-01 -1.05938232e+00 -3.29409212e-01
-7.20811561e-02 2.41187602e-01 5.14955580e-01 6.17809355e-01
3.38166445e-01 4.85739827e-01 9.58357379e-02 -7.09939063e-01
-6.52253091e-01 -1.44094276e+00 -5.08096039e-01 2.17147946e-01
1.30106449e-01 3.98637466e-02 -1.55097976e-01 3.90014529e-01] | [11.649471282958984, 9.019083023071289] |
2a98e961-d4d5-4beb-bda9-b21dd10d9da6 | efficient-palm-line-segmentation-with-u-net | 2102.12127 | null | https://arxiv.org/abs/2102.12127v1 | https://arxiv.org/pdf/2102.12127v1.pdf | Efficient Palm-Line Segmentation with U-Net Context Fusion Module | Many cultures around the world believe that palm reading can be used to predict the future life of a person. Palmistry uses features of the hand such as palm lines, hand shape, or fingertip position. However, the research on palm-line detection is still scarce, many of them applied traditional image processing techniques. In most real-world scenarios, images usually are not in well-conditioned, causing these methods to severely under-perform. In this paper, we propose an algorithm to extract principle palm lines from an image of a person's hand. Our method applies deep learning networks (DNNs) to improve performance. Another challenge of this problem is the lack of training data. To deal with this issue, we handcrafted a dataset from scratch. From this dataset, we compare the performance of readily available methods with ours. Furthermore, based on the UNet segmentation neural network architecture and the knowledge of attention mechanism, we propose a highly efficient architecture to detect palm-lines. We proposed the Context Fusion Module to capture the most important context feature, which aims to improve segmentation accuracy. The experimental results show that it outperforms the other methods with the highest F1 Score about 99.42% and mIoU is 0.584 for the same dataset. | ['Ta Minh Thanh', 'Ngoc N. Tran', 'Linh Bao Doan', 'Son Trung Nguyen', 'Toan Pham Van'] | 2021-02-24 | null | null | null | null | ['unet-segmentation', 'line-detection'] | ['computer-vision', 'computer-vision'] | [ 2.02206179e-01 -4.80648965e-01 -1.11760475e-01 -3.06274652e-01
-6.91062585e-02 -5.85130572e-01 2.36178428e-01 -5.13955891e-01
-4.12947208e-01 4.59060848e-01 -2.59851068e-01 -4.72779348e-02
6.83174729e-02 -7.66989768e-01 -4.66541439e-01 -5.58534026e-01
3.89184713e-01 9.81076583e-02 1.83762491e-01 5.99383861e-02
3.70645791e-01 6.82084441e-01 -1.30495346e+00 1.38562217e-01
9.02566612e-01 1.00874925e+00 1.73910335e-01 5.41257679e-01
-1.78463370e-01 2.19750032e-01 -5.70016086e-01 -6.85773492e-01
3.26474726e-01 -3.40775609e-01 -6.71958208e-01 5.59325479e-02
5.90429485e-01 -7.70736575e-01 -3.99001628e-01 1.20357800e+00
7.57316709e-01 -1.33826926e-01 3.62030596e-01 -1.02644432e+00
-6.38431072e-01 4.08013046e-01 -9.48001027e-01 3.41059035e-03
3.18308890e-01 6.81615993e-02 5.86090803e-01 -6.15136206e-01
4.03139740e-01 1.25677550e+00 8.80116999e-01 6.17647827e-01
-4.95057374e-01 -7.11382866e-01 1.91899195e-01 3.34645778e-01
-1.45346320e+00 -1.68054745e-01 1.05232823e+00 -6.14735894e-02
3.16486388e-01 3.67071122e-01 9.99890924e-01 1.21364689e+00
1.37220651e-01 1.10486102e+00 1.53399110e+00 -6.68354392e-01
-1.50728926e-01 1.51489556e-01 2.68847048e-01 7.86315203e-01
1.90011308e-01 -1.72279313e-01 -3.44580531e-01 1.62516773e-01
1.24985969e+00 4.39550161e-01 -1.55691817e-01 1.79200113e-01
-1.02934253e+00 3.66430879e-01 4.48630929e-01 5.06908119e-01
-4.55654830e-01 -1.36244014e-01 -1.67134687e-01 -2.14440897e-01
-2.52720267e-02 -2.25185100e-02 -2.93327153e-01 -5.81929795e-02
-1.12451077e+00 3.53369594e-01 8.73233080e-01 8.28461111e-01
3.98732811e-01 -4.15018409e-01 -1.63277492e-01 7.12822497e-01
3.98842722e-01 6.73610270e-01 3.74746680e-01 -8.18374872e-01
2.30951622e-01 6.87692702e-01 1.01860784e-01 -1.18510449e+00
-4.70456511e-01 -3.37648809e-01 -7.93812573e-01 1.36117414e-01
7.80545592e-01 -4.05810446e-01 -1.10776985e+00 1.30103660e+00
2.21188277e-01 7.05352277e-02 -3.96872431e-01 1.21053529e+00
7.34024882e-01 4.78248954e-01 -1.54904291e-01 -2.79424023e-02
1.44020450e+00 -1.04519677e+00 -8.65264535e-01 -2.49584064e-01
-1.76751688e-01 -9.61855769e-01 1.01155627e+00 8.33536208e-01
-6.73032880e-01 -7.45483160e-01 -8.32118571e-01 1.88595816e-01
-4.22157288e-01 4.00112838e-01 7.88124502e-01 1.02847207e+00
-6.91649914e-01 6.04463935e-01 -6.69616461e-01 -7.59083867e-01
5.38265347e-01 3.10231060e-01 -6.89282119e-02 3.36297937e-02
-9.68044162e-01 5.23400009e-01 4.70819682e-01 5.59495330e-01
-4.59476680e-01 -8.91621709e-02 -1.85519814e-01 -2.04511151e-01
4.92016524e-01 -4.71809655e-01 1.07278097e+00 -1.29977429e+00
-1.61273432e+00 5.53785324e-01 -2.62194693e-01 -1.30575985e-01
7.82860816e-01 -8.12443316e-01 -5.58677435e-01 3.26337069e-01
-2.51795620e-01 5.72244763e-01 1.12038159e+00 -1.30040884e+00
-8.22374821e-01 -8.15691233e-01 -7.43660852e-02 1.86498567e-01
-5.72002351e-01 5.29527329e-02 -9.33833897e-01 -6.88197017e-01
1.51066789e-02 -9.71632004e-01 1.71694458e-02 7.95045272e-02
-6.91879272e-01 -4.21077788e-01 1.22566652e+00 -1.29307151e+00
1.12871909e+00 -1.78471935e+00 -3.41884047e-01 3.85598838e-01
6.77285716e-02 6.92080140e-01 8.74434784e-02 2.80690879e-01
2.78622895e-01 9.10171792e-02 -2.17364207e-01 1.98595971e-02
-1.76035821e-01 1.87630113e-02 -8.43384396e-03 3.09134185e-01
-2.14668605e-02 8.08754146e-01 -4.63524401e-01 -8.06938708e-01
2.68012941e-01 7.42479563e-01 -2.13152394e-01 3.87970239e-01
-1.18879408e-01 4.62259769e-01 -4.95058358e-01 8.99408638e-01
7.85394788e-01 -1.74284689e-02 1.51001692e-01 -3.65477324e-01
-1.91365331e-02 -4.58977044e-01 -1.36417413e+00 1.91466463e+00
-1.67120740e-01 5.34726441e-01 -7.97004327e-02 -4.90555733e-01
9.25191045e-01 3.00596654e-01 4.28466856e-01 -5.20055234e-01
2.77178168e-01 -7.18442872e-02 1.98818725e-02 -9.31433856e-01
1.36830613e-01 2.40510747e-01 5.00161469e-01 3.60416502e-01
-1.49657935e-01 3.40636253e-01 -1.15197161e-02 -1.24256536e-01
5.04741073e-01 5.11891186e-01 -3.34863551e-02 6.28105029e-02
5.31664968e-01 -2.52141148e-01 5.19504428e-01 7.66135037e-01
-3.21266830e-01 7.73222506e-01 2.43108377e-01 -5.24755418e-01
-7.92413354e-01 -6.90213799e-01 3.27515863e-02 1.19397390e+00
3.53646517e-01 -2.17908159e-01 -1.34553719e+00 -8.00054669e-01
-2.31731802e-01 3.84003401e-01 -4.61256832e-01 3.05871069e-01
-7.18328238e-01 -8.87812555e-01 5.90573251e-01 7.73028135e-01
1.14213645e+00 -1.14242554e+00 -6.76543355e-01 -5.78316599e-02
-1.45877212e-01 -1.21895814e+00 -4.65430260e-01 -4.24909055e-01
-7.67248750e-01 -1.10525060e+00 -1.27016461e+00 -9.46612716e-01
5.65344572e-01 1.57517776e-01 6.06481016e-01 -4.09692712e-02
-4.62222487e-01 3.18187267e-01 -3.23035449e-01 -6.48517668e-01
2.94214189e-01 1.89619422e-01 -7.46844113e-02 2.91773885e-01
6.72109604e-01 -4.27635312e-01 -8.58804464e-01 2.01644450e-01
-6.12533092e-01 1.14143267e-01 1.07950652e+00 5.54048538e-01
4.49452877e-01 1.91520959e-01 4.17309195e-01 -8.41133535e-01
8.06452811e-01 8.15883577e-02 -2.66973287e-01 4.11402017e-01
-6.01844251e-01 -3.28126997e-01 4.63535100e-01 -4.46165919e-01
-1.27752364e+00 3.01086187e-01 -2.30122477e-01 -2.71477729e-01
-5.91499388e-01 8.64012539e-02 -3.95786375e-01 -5.55610880e-02
2.04260334e-01 3.44237387e-01 -4.80726734e-02 -7.12803185e-01
3.69380355e-01 1.15580797e+00 8.08935642e-01 -5.03830671e-01
5.43747544e-01 5.23242235e-01 -2.55482346e-01 -1.03506660e+00
-8.31045926e-01 -4.00164336e-01 -1.02058232e+00 -3.16478550e-01
8.56746733e-01 -3.58454555e-01 -9.66396570e-01 9.32037354e-01
-1.18144310e+00 -5.08769089e-03 3.01628470e-01 2.85648376e-01
-2.04533353e-01 7.64876485e-01 -4.60079521e-01 -1.21846354e+00
-7.90550590e-01 -8.68547082e-01 8.67105186e-01 9.11492884e-01
2.73537412e-02 -7.95195282e-01 -2.59347469e-01 4.94587809e-01
2.02896923e-01 3.02211195e-01 5.09581149e-01 -6.75909698e-01
-2.98978835e-01 -3.87389272e-01 -4.83875394e-01 4.20617670e-01
2.25901201e-01 1.21167764e-01 -1.24503183e+00 -2.06708565e-01
-4.69240025e-02 3.76671292e-02 7.73219526e-01 2.44392693e-01
1.70564330e+00 -2.54005134e-01 -4.99380082e-01 5.67956507e-01
1.29029715e+00 2.52448082e-01 5.08067846e-01 3.45105797e-01
1.04265273e+00 7.34262168e-01 4.52460587e-01 4.77311611e-01
2.51990139e-01 3.30662161e-01 2.88551927e-01 -9.17534977e-02
-2.77708769e-01 -2.26171628e-01 2.02107444e-01 5.11675715e-01
-6.47615910e-01 -1.93957835e-01 -9.41877186e-01 3.69798750e-01
-1.76059926e+00 -8.29663157e-01 -1.51805997e-01 1.93600488e+00
7.01546192e-01 1.52236253e-01 5.17069161e-01 2.36041516e-01
1.05308282e+00 -7.95915127e-02 -6.07050002e-01 -5.77223487e-02
1.64314788e-02 1.80123672e-01 3.61088485e-01 1.36832535e-01
-1.35938323e+00 1.03439963e+00 6.11103392e+00 6.80941939e-01
-1.29186189e+00 -2.19417810e-01 3.61016899e-01 2.23620951e-01
1.68327361e-01 -3.54061455e-01 -9.30752993e-01 6.89139843e-01
5.09550869e-01 4.48721528e-01 5.17195106e-01 5.27437747e-01
5.24211749e-02 9.81158018e-02 -9.32039618e-01 1.44389379e+00
3.53448898e-01 -5.91674626e-01 6.91047460e-02 -9.72567275e-02
4.26464498e-01 -5.20181000e-01 2.09618166e-01 -2.00313270e-01
-2.13318735e-01 -1.23233247e+00 5.28473973e-01 8.78066242e-01
6.29282475e-01 -9.54315245e-01 8.10469389e-01 5.12134075e-01
-8.59475315e-01 -6.63306937e-02 -1.97244316e-01 1.92950796e-02
-1.86139211e-01 5.15547574e-01 -1.00400722e+00 4.57935482e-01
9.17377830e-01 5.04596531e-01 -7.42582500e-01 1.10943890e+00
-4.82632875e-01 6.90033078e-01 -5.29091775e-01 -2.08343148e-01
1.33913085e-01 -1.61255300e-01 3.10075790e-01 1.36676717e+00
2.26515666e-01 3.16990949e-02 5.91837727e-02 9.47403371e-01
3.65977734e-02 2.02440843e-01 -3.01080614e-01 -1.60898224e-01
3.58342767e-01 1.49885976e+00 -8.07280481e-01 -6.16698489e-02
-4.97080237e-01 1.38569689e+00 -1.80945575e-01 4.77261871e-01
-6.43774569e-01 -7.22094297e-01 1.18047021e-01 1.31501794e-01
2.03610346e-01 -1.48561940e-01 -3.68789852e-01 -1.00251997e+00
3.01907599e-01 -8.28513443e-01 2.49392197e-01 -5.74790299e-01
-1.32027900e+00 4.18539822e-01 -3.51148397e-01 -6.73158407e-01
-1.14185527e-01 -9.59786236e-01 -7.60850549e-01 9.21741843e-01
-1.46095455e+00 -1.75616002e+00 -7.28414714e-01 8.14772844e-01
6.06287897e-01 -2.11486310e-01 7.07084537e-01 1.59520999e-01
-8.54993165e-01 7.82141149e-01 4.23986837e-02 7.68593788e-01
9.16816831e-01 -1.29572761e+00 3.79162878e-01 9.47958231e-01
1.21072121e-01 8.35983217e-01 3.31609637e-01 -7.23632991e-01
-1.52422893e+00 -8.44201267e-01 3.83773685e-01 -3.02379936e-01
1.03414573e-01 -8.75501111e-02 -7.26100862e-01 6.30126536e-01
3.04035038e-01 -9.33477283e-02 7.89067805e-01 1.94586679e-01
-1.31741270e-01 -3.59440714e-01 -1.20370054e+00 5.06912231e-01
9.90278602e-01 -2.38621548e-01 -7.04324543e-01 5.22402301e-02
4.46358547e-02 -2.36216739e-01 -8.24752331e-01 3.16487998e-01
1.23716259e+00 -1.05418956e+00 9.96191919e-01 -3.18765521e-01
1.78437397e-01 -6.11483492e-02 1.25118777e-01 -9.02469993e-01
-2.11550012e-01 -5.58393538e-01 -4.04235482e-01 1.59769082e+00
-1.87257677e-01 -5.43739736e-01 9.97982144e-01 4.85915005e-01
3.62317353e-01 -8.17129314e-01 -3.19635302e-01 -4.88601744e-01
1.10230520e-02 -2.48769403e-01 6.93283558e-01 6.99192464e-01
-3.74820471e-01 1.49338678e-01 -4.66218621e-01 2.46984676e-01
9.61296737e-01 4.52092201e-01 7.39265978e-01 -1.47116137e+00
-1.26084268e-01 -3.83676171e-01 -1.42323568e-01 -1.10533726e+00
-1.53360695e-01 -4.45835382e-01 -2.05693051e-01 -1.59077740e+00
4.28311110e-01 -2.32070148e-01 -3.80683362e-01 5.11980176e-01
-4.45236444e-01 2.78530061e-01 3.62247080e-01 2.02405453e-01
-4.73959148e-01 -4.31409143e-02 1.51657128e+00 -1.58149824e-01
-2.56727189e-01 2.59499699e-01 -7.42531598e-01 1.28954005e+00
1.02040064e+00 5.89713603e-02 -5.37112616e-02 -5.59585690e-01
-2.23590866e-01 -1.71118841e-01 4.85399306e-01 -1.18162739e+00
3.91124427e-01 -2.12949261e-01 1.22131169e+00 -9.02885437e-01
2.68798530e-01 -9.57019925e-01 -3.27882290e-01 4.44870949e-01
-1.07122762e-02 -3.09966743e-01 -9.84961167e-02 3.98807675e-01
1.71180397e-01 -3.81213158e-01 5.78065217e-01 -2.70599902e-01
-1.14435196e+00 3.77994090e-01 -1.90754682e-02 -3.99872512e-01
9.45678294e-01 -3.91792148e-01 -3.43038514e-02 -1.68344364e-01
-6.03372335e-01 2.78800651e-02 3.20471674e-01 5.03502786e-01
7.15709031e-01 -1.20287597e+00 -6.74768031e-01 3.27117741e-01
-3.03383976e-01 1.04493886e-01 -2.22068764e-02 5.01617491e-01
-7.82594502e-01 3.91928434e-01 -6.62470520e-01 -4.96514797e-01
-1.35890365e+00 4.17482585e-01 3.43304008e-01 1.96951553e-01
-6.56973541e-01 7.44445682e-01 -1.91107064e-01 -1.87495977e-01
5.05950451e-01 -1.20712049e-01 -4.72625196e-01 -1.08669614e-02
7.43467510e-01 6.47614181e-01 -2.14813948e-01 -4.79251862e-01
-3.01548302e-01 9.68283892e-01 -1.93203285e-01 -8.33475441e-02
1.16655648e+00 2.14555133e-02 -2.11568341e-01 2.23735467e-01
8.07170510e-01 1.55259669e-02 -1.19325149e+00 -8.14409032e-02
-2.40997195e-01 -6.53068364e-01 -5.45333438e-02 -1.14224803e+00
-1.17474461e+00 1.22878444e+00 1.13805139e+00 1.02558285e-01
1.25410056e+00 -3.15691859e-01 1.23516190e+00 4.97055233e-01
3.74923885e-01 -1.27275145e+00 -1.64107271e-02 1.83423385e-01
6.54656112e-01 -1.29630983e+00 -6.24073111e-02 -4.60032493e-01
-3.89408380e-01 1.50338972e+00 7.41349220e-01 -1.19490534e-01
4.55591470e-01 1.41040176e-01 2.34149903e-01 9.63985845e-02
2.43793815e-01 -2.46371463e-01 4.17046428e-01 7.80202568e-01
4.86688703e-01 1.40530154e-01 -2.92697996e-01 1.03735769e+00
-2.70622790e-01 3.37279052e-01 3.85124758e-02 9.30359185e-01
-4.82409537e-01 -1.10244846e+00 -8.80677879e-01 4.93732005e-01
-8.49361897e-01 2.84332901e-01 -7.39039660e-01 6.51424229e-01
5.37376463e-01 8.76621902e-01 -1.29382282e-01 -4.14244026e-01
1.82685062e-01 1.98728934e-01 8.11803162e-01 -1.91341251e-01
-3.68816167e-01 2.90889621e-01 -4.64186281e-01 -3.33190501e-01
-4.97590750e-01 -5.34020543e-01 -1.12596142e+00 -3.53227943e-01
-3.56548309e-01 -2.65252292e-01 5.87631166e-01 9.78305697e-01
-6.80487603e-02 4.66791809e-01 3.19339275e-01 -7.05263853e-01
-4.19221312e-01 -1.04238641e+00 -6.92994714e-01 5.44792295e-01
6.88157603e-02 -3.70364994e-01 3.33989173e-01 2.92201847e-01] | [6.5646185874938965, -0.5085762143135071] |
8b1a3500-9760-4067-a460-23add902f8f2 | dcso-dynamic-combination-of-detector-scores | 1911.10418 | null | https://arxiv.org/abs/1911.10418v1 | https://arxiv.org/pdf/1911.10418v1.pdf | DCSO: Dynamic Combination of Detector Scores for Outlier Ensembles | Selecting and combining the outlier scores of different base detectors used within outlier ensembles can be quite challenging in the absence of ground truth. In this paper, an unsupervised outlier detector combination framework called DCSO is proposed, demonstrated and assessed for the dynamic selection of most competent base detectors, with an emphasis on data locality. The proposed DCSO framework first defines the local region of a test instance by its k nearest neighbors and then identifies the top-performing base detectors within the local region. Experimental results on ten benchmark datasets demonstrate that DCSO provides consistent performance improvement over existing static combination approaches in mining outlying objects. To facilitate interpretability and reliability of the proposed method, DCSO is analyzed using both theoretical frameworks and visualization techniques, and presented alongside empirical parameter setting instructions that can be used to improve the overall performance. | ['Yue Zhao', 'Maciej K. Hryniewicki'] | 2019-11-23 | null | null | null | null | ['outlier-ensembles'] | ['methodology'] | [-1.92640662e-01 -3.43548328e-01 1.12065189e-01 -3.02676912e-02
-7.55421221e-01 -4.50798452e-01 3.71309131e-01 9.48447645e-01
-1.58444226e-01 4.04083967e-01 5.00412099e-02 -9.60924104e-02
-8.85239661e-01 -4.27041471e-01 -1.22130863e-01 -7.78649211e-01
-5.72195530e-01 4.47753996e-01 3.55588347e-01 1.61322936e-01
7.60029495e-01 7.62658179e-01 -1.61958265e+00 2.24133030e-01
1.19709682e+00 7.83849180e-01 -3.13716441e-01 3.18300545e-01
1.94838196e-01 4.14032161e-01 -1.00328338e+00 1.68835593e-03
6.58241272e-01 -2.92325944e-01 -3.41105424e-02 -1.41530886e-01
5.23155749e-01 -2.05585271e-01 7.45165255e-03 8.85953367e-01
5.85093737e-01 2.55024314e-01 7.69588709e-01 -1.37317336e+00
-2.23171115e-01 6.09046996e-01 -4.17412311e-01 8.49379838e-01
3.48945916e-01 3.00287932e-01 8.71854007e-01 -1.22122395e+00
4.35019165e-01 7.93967366e-01 6.84122860e-01 -1.06324024e-01
-1.13478518e+00 -5.88219523e-01 1.51795506e-01 2.04576805e-01
-1.86097193e+00 -1.90252706e-01 5.87922752e-01 -3.50270718e-01
1.04539561e+00 4.68858272e-01 4.20057178e-01 6.81937277e-01
2.68582463e-01 4.15849954e-01 1.04771471e+00 -4.03498024e-01
6.48777068e-01 -1.14792071e-01 4.75406706e-01 3.90134960e-01
1.18841422e+00 2.05582798e-01 -8.51318181e-01 -8.12813997e-01
1.74491212e-01 2.08380982e-01 -7.99689144e-02 -4.31505531e-01
-8.94616723e-01 5.72018743e-01 3.93382609e-01 1.59443051e-01
-4.54406589e-01 -1.43624187e-01 5.07388711e-01 3.50500762e-01
3.00033569e-01 5.76638699e-01 -7.80466944e-02 -3.29484157e-02
-1.16547298e+00 3.30068231e-01 5.99070132e-01 9.47337747e-01
3.60560834e-01 5.18315472e-02 -3.80484968e-01 6.14461064e-01
3.27896416e-01 1.85687765e-01 3.37082654e-01 -3.71662736e-01
6.31690621e-01 1.23212016e+00 2.70041645e-01 -1.18555653e+00
-3.99523765e-01 -5.50840855e-01 -4.98039961e-01 4.07405019e-01
1.62915304e-01 5.97802624e-02 -9.42773283e-01 9.66934681e-01
4.12059724e-01 4.90546286e-01 1.77901953e-01 6.35782838e-01
5.31055570e-01 2.26463601e-01 5.63628972e-02 -3.84817481e-01
8.44163120e-01 -2.94402897e-01 -3.58616859e-01 1.45679280e-01
8.47909391e-01 -5.59293687e-01 5.83304167e-01 5.23487329e-01
-6.56439841e-01 -2.44880244e-01 -1.18634486e+00 5.70962489e-01
-3.91978443e-01 2.38024428e-01 2.32559353e-01 6.11626267e-01
-7.87177145e-01 5.99696457e-01 -9.45064366e-01 -5.44058502e-01
4.30953234e-01 5.65474570e-01 -1.34610981e-01 -8.43851939e-02
-7.54482746e-01 6.11567259e-01 8.83443892e-01 1.03125878e-01
-7.75903761e-01 -6.20624125e-01 -5.38200259e-01 1.32474691e-01
4.61876899e-01 -5.51911831e-01 5.56496501e-01 -4.33223695e-01
-7.41488218e-01 3.38414103e-01 -6.12008274e-02 -7.32810378e-01
7.36384153e-01 -4.72452193e-01 -6.17012382e-01 6.25355169e-02
1.67481244e-01 -1.42798066e-01 7.26072192e-01 -1.23514760e+00
-1.03832889e+00 -5.67144811e-01 -3.59484613e-01 3.33903491e-01
-1.88958630e-01 5.99453002e-02 -4.02776867e-01 -6.86967194e-01
6.30026281e-01 -6.95493639e-01 -3.69108647e-01 -4.91012424e-01
-6.41870499e-01 -3.17313850e-01 1.03420985e+00 -1.31679460e-01
1.95177090e+00 -2.29830194e+00 -2.51723200e-01 1.04593384e+00
3.40506285e-01 2.21574321e-01 2.35710904e-01 7.54874766e-01
-2.62899343e-02 9.23715830e-02 -1.83900416e-01 -1.07489690e-01
-1.57832667e-01 1.24730216e-02 -2.14786500e-01 5.96467614e-01
1.20570228e-01 2.26857916e-01 -9.34396088e-01 -3.16957682e-01
3.52850676e-01 8.55407268e-02 -3.92951608e-01 1.25135645e-01
3.59733969e-01 3.00679684e-01 -4.61281925e-01 1.14279306e+00
6.49336755e-01 1.65658399e-01 4.90039401e-02 1.23360315e-02
4.43767868e-02 -1.48309961e-01 -1.83059144e+00 7.50173986e-01
1.36783347e-01 2.12600067e-01 -4.80758727e-01 -7.18349576e-01
1.47234261e+00 7.55010694e-02 6.51529074e-01 -4.93887663e-01
-1.24246016e-01 6.85980558e-01 2.36554965e-01 -4.12009716e-01
4.14054245e-01 4.07836050e-01 -2.08900899e-01 3.39542001e-01
-2.21798584e-01 4.59122926e-01 6.33453965e-01 2.68678069e-01
1.54820967e+00 -2.88115025e-01 8.11001778e-01 -3.95972759e-01
4.31485862e-01 1.07087530e-01 7.26220548e-01 1.25067937e+00
-4.55640048e-01 5.77692807e-01 4.75599766e-01 -5.12167513e-01
-1.07567751e+00 -1.36888778e+00 -4.51222062e-01 7.03444719e-01
3.01917762e-01 -6.37275815e-01 -4.51316744e-01 -7.25375414e-01
5.80979049e-01 7.70242393e-01 -5.60490668e-01 -1.49017394e-01
-4.17897493e-01 -9.07112718e-01 6.54124081e-01 4.49249178e-01
1.56534448e-01 -9.60450947e-01 -8.56978357e-01 5.70153594e-02
4.65106547e-01 -7.51971602e-01 -1.28005609e-01 5.20227730e-01
-1.19643354e+00 -1.30061412e+00 -4.13151383e-02 -2.94625461e-01
1.00815237e+00 2.88386226e-01 1.01147521e+00 4.80815530e-01
-2.49326482e-01 1.96761325e-01 -5.90800703e-01 -5.13716936e-01
-1.60846189e-01 -9.74364132e-02 4.23534513e-01 -7.84879327e-02
7.93921411e-01 -5.53421199e-01 -5.78957260e-01 3.44468921e-01
-9.04528677e-01 -7.61705101e-01 5.60803831e-01 8.22600722e-01
7.57097781e-01 1.64810345e-01 6.28992558e-01 -1.13613498e+00
9.64104414e-01 -8.67949009e-01 -6.62371337e-01 1.42347246e-01
-9.52556551e-01 -6.83759823e-02 6.03208482e-01 3.65058556e-02
-5.99241674e-01 7.07080774e-03 5.19847512e-01 -4.01751220e-01
-3.95984292e-01 5.41148782e-01 -1.89424176e-02 1.33110195e-01
1.02110457e+00 -6.64341748e-02 -5.10107100e-01 -4.24068272e-01
-1.06380843e-01 6.72917783e-01 5.76115012e-01 -6.01746023e-01
8.14059258e-01 4.50332016e-01 7.01048151e-02 -5.53554296e-01
-5.29035568e-01 -1.04240048e+00 -1.01200700e+00 -1.77988559e-01
4.05352563e-02 -7.65872538e-01 -1.91627890e-01 1.31468251e-01
-5.09435177e-01 3.15648943e-01 -2.66917348e-01 2.70629019e-01
-2.41916209e-01 3.13118011e-01 -7.02206716e-02 -9.65834796e-01
-3.63786250e-01 -1.04374874e+00 9.29987192e-01 1.85524479e-01
-4.33665127e-01 -8.17474246e-01 1.90073088e-01 1.88309386e-01
1.75427496e-02 7.25316644e-01 6.48310125e-01 -1.49100184e+00
-6.24839842e-01 -6.75788760e-01 3.17470133e-02 3.99455614e-02
1.65577129e-01 3.75796765e-01 -7.44500816e-01 -5.83095789e-01
-3.71147722e-01 2.78960884e-01 7.08034992e-01 1.76096648e-01
9.14240956e-01 -1.15164384e-01 -6.66359663e-01 6.53272748e-01
1.79945683e+00 2.02552080e-01 4.43397462e-01 7.02060342e-01
5.31973839e-01 9.51043591e-02 9.47474718e-01 9.33791280e-01
-2.47252986e-01 5.08971870e-01 4.92645442e-01 1.21008717e-01
3.57407987e-01 -3.60521898e-02 4.43632424e-01 5.60977578e-01
-9.47540849e-02 -1.69721767e-01 -1.31941426e+00 7.17148542e-01
-1.98088384e+00 -1.06798065e+00 -2.68713951e-01 2.67053175e+00
9.10318829e-03 4.83625174e-01 3.12291682e-01 3.53482485e-01
8.14449728e-01 -2.16111183e-01 -6.57967091e-01 -5.03659189e-01
-1.08239420e-01 1.37257710e-01 5.31602859e-01 2.28321433e-01
-1.07195306e+00 4.74350780e-01 6.66262054e+00 5.14146030e-01
-7.39165306e-01 -1.61268458e-01 3.37130159e-01 -4.14088368e-01
1.80516064e-01 6.87656701e-02 -7.96959758e-01 6.56389475e-01
1.02558529e+00 -4.66613263e-01 -1.41799897e-01 8.96325707e-01
6.49724364e-01 -4.60265458e-01 -1.14205945e+00 9.46583807e-01
5.90948798e-02 -1.07918358e+00 9.19867828e-02 1.03101574e-01
1.01324487e+00 1.66635782e-01 -3.52696851e-02 7.05760717e-02
2.83622593e-01 -8.64717841e-01 5.17219424e-01 5.24626732e-01
1.90813452e-01 -1.18420255e+00 1.17762756e+00 1.61245957e-01
-1.13244557e+00 -7.73062408e-01 -2.79629380e-01 -1.96603220e-03
-3.13621670e-01 7.61512101e-01 -1.15342689e+00 9.23964977e-01
9.65000153e-01 5.47853470e-01 -1.14659488e+00 1.73545921e+00
6.00546077e-02 8.09745669e-01 -6.92162812e-01 5.90887442e-02
1.13997206e-01 -1.91093996e-01 8.84525537e-01 1.30470300e+00
5.75171590e-01 -1.99124739e-01 4.06709284e-01 4.77054864e-01
2.49950528e-01 5.41967273e-01 -8.96243930e-01 4.13161188e-01
1.07220387e+00 8.96237493e-01 -1.13670695e+00 -3.14999491e-01
-2.22429886e-01 4.90230024e-01 2.91241169e-01 3.89844537e-01
-7.20109820e-01 -3.77312005e-01 5.45336604e-01 2.32862383e-01
1.92052692e-01 -7.55092651e-02 -8.37220371e-01 -8.27961326e-01
3.66206825e-01 -8.80580366e-01 9.86724257e-01 -1.39381096e-01
-1.27545285e+00 3.56110185e-01 1.78261757e-01 -1.88469529e+00
3.59597988e-02 -5.37848175e-01 -1.03628278e+00 6.04283035e-01
-6.99797750e-01 -5.59575140e-01 -4.96985674e-01 2.10605502e-01
2.22825706e-01 -3.54076326e-01 5.15729606e-01 1.61012858e-01
-9.56182241e-01 7.48980701e-01 7.33834743e-01 -1.28328383e-01
7.02338815e-01 -1.36865640e+00 1.81577221e-01 1.38973892e+00
1.66261017e-01 9.07693803e-01 8.47801805e-01 -9.15825784e-01
-8.95860970e-01 -1.19142485e+00 4.08758968e-01 -6.94964945e-01
5.39018750e-01 6.11588955e-02 -1.08618462e+00 3.90188098e-01
-3.22782159e-01 2.36968040e-01 8.69029522e-01 2.52159268e-01
-1.50577828e-01 -2.54158556e-01 -1.34713614e+00 4.97498691e-01
9.05291796e-01 -1.95817739e-01 -5.93205631e-01 1.45634308e-01
-8.84535089e-02 -4.27914441e-01 -8.68471682e-01 4.94263351e-01
1.34476990e-01 -1.35875750e+00 7.14416862e-01 -5.23643851e-01
-2.41920620e-01 -9.28494453e-01 -1.64749339e-01 -1.08285403e+00
-2.04990745e-01 -4.43213224e-01 -2.86078125e-01 1.17322612e+00
4.34100598e-01 -5.54291546e-01 6.91949487e-01 3.02002460e-01
-2.18772456e-01 -9.99176860e-01 -1.05534196e+00 -8.80534291e-01
-3.58779848e-01 -4.49323654e-01 5.95063627e-01 6.92676783e-01
1.03729730e-02 -1.99129358e-01 -3.99107970e-02 8.40792835e-01
6.79327965e-01 -1.33499607e-01 8.47334325e-01 -1.47485900e+00
8.78295451e-02 -3.51572037e-01 -1.04680312e+00 -1.71800688e-01
-5.20590186e-01 -9.08744454e-01 -2.10995078e-01 -1.23301518e+00
1.14175230e-01 -4.77482766e-01 -9.79947507e-01 2.50670731e-01
-5.80891132e-01 3.00157309e-01 4.27467264e-02 6.87411129e-01
-8.82674217e-01 1.34254351e-01 2.74122864e-01 1.21207170e-01
-5.95215321e-01 -6.88546477e-03 -4.60248858e-01 7.24600077e-01
7.54966736e-01 -8.32072258e-01 -3.98001254e-01 1.12664618e-01
3.70535441e-02 -4.82909679e-01 2.40563527e-01 -1.55038977e+00
3.43805015e-01 -1.93197131e-01 5.82598388e-01 -8.53775799e-01
-3.76949459e-01 -1.00470185e+00 2.54911274e-01 5.87274194e-01
-9.07249004e-02 6.23225033e-01 8.50140080e-02 7.91691840e-01
-3.25667381e-01 -3.56424779e-01 7.74762213e-01 2.48693317e-01
-8.95449638e-01 -3.59824076e-02 -1.45620350e-02 -8.21634531e-02
1.53247762e+00 -6.85095191e-01 -5.25284633e-02 1.29875302e-01
-7.87019432e-01 4.26764816e-01 7.18784869e-01 1.87984914e-01
8.45660031e-01 -1.33551860e+00 -7.30723143e-01 1.74325436e-01
7.58080363e-01 -3.31444964e-02 1.66602969e-01 8.52114260e-01
-8.25590789e-01 3.41395736e-01 -1.41049623e-01 -8.19710135e-01
-1.40851581e+00 5.16540885e-01 3.60860467e-01 -3.03845137e-01
-8.45487058e-01 2.94313371e-01 -4.92747575e-01 -2.40703020e-02
3.58415872e-01 -3.36878628e-01 -1.89997345e-01 4.73155864e-02
4.62250501e-01 8.91368628e-01 4.35354233e-01 -5.44413984e-01
-5.98681688e-01 1.68985277e-01 -1.24502867e-01 3.56353819e-01
1.07990336e+00 -7.87664875e-02 -3.98322083e-02 4.88412380e-01
5.45176506e-01 2.82325923e-01 -9.73938942e-01 4.94226255e-02
8.10788155e-01 -7.02751517e-01 -3.53181124e-01 -6.50162101e-01
-6.16997957e-01 2.32887700e-01 9.09095705e-01 2.34766826e-01
1.31592405e+00 -2.43628845e-01 1.39877066e-01 5.10177493e-01
3.46217245e-01 -1.24631894e+00 -5.85921519e-02 2.50563592e-01
4.92104858e-01 -1.24563694e+00 4.20386165e-01 -2.46983781e-01
-4.85775590e-01 1.19287884e+00 8.08608711e-01 -5.33453047e-01
5.90108633e-01 2.56961256e-01 9.46213603e-02 -4.17073876e-01
-8.50274026e-01 -4.44819741e-02 3.72290999e-01 6.00766361e-01
3.21614295e-01 4.15895544e-02 -6.04170978e-01 2.82637656e-01
5.46901897e-02 -1.96764559e-01 4.78105068e-01 9.72393334e-01
-6.37661755e-01 -8.34869564e-01 -8.02124977e-01 1.04995716e+00
-4.28116351e-01 3.43809985e-02 -5.01130402e-01 8.63303602e-01
5.01652241e-01 8.23760152e-01 1.52995661e-01 -2.98965752e-01
5.89860499e-01 1.88272342e-01 -1.78232342e-01 -6.72348142e-01
-7.76792228e-01 -1.60555899e-01 -1.00567922e-01 -4.66528088e-01
-3.01169250e-02 -8.33080053e-01 -1.22964096e+00 -6.60716593e-02
-5.04008710e-01 3.46191347e-01 1.44887567e-01 8.92168045e-01
6.40997827e-01 4.67770487e-01 6.56600475e-01 -4.16480273e-01
-5.43491900e-01 -8.78580034e-01 -5.87760091e-01 7.87915945e-01
2.37782210e-01 -8.69772792e-01 -6.91487849e-01 -2.89787918e-01] | [7.543262481689453, 2.736804962158203] |
a78b2376-d5da-4b25-be50-785389bd0b41 | bora-bayesian-optimization-for-resource | 2210.05977 | null | https://arxiv.org/abs/2210.05977v1 | https://arxiv.org/pdf/2210.05977v1.pdf | BORA: Bayesian Optimization for Resource Allocation | Optimal resource allocation is gaining a renewed interest due its relevance as a core problem in managing, over time, cloud and high-performance computing facilities. Semi-Bandit Feedback (SBF) is the reference method for efficiently solving this problem. In this paper we propose (i) an extension of the optimal resource allocation to a more general class of problems, specifically with resources availability changing over time, and (ii) Bayesian Optimization as a more efficient alternative to SBF. Three algorithms for Bayesian Optimization for Resource Allocation, namely BORA, are presented, working on allocation decisions represented as numerical vectors or distributions. The second option required to consider the Wasserstein distance as a more suitable metric to use into one of the BORA algorithms. Results on (i) the original SBF case study proposed in the literature, and (ii) a real-life application (i.e., the optimization of multi-channel marketing) empirically prove that BORA is a more efficient and effective learning-and-optimization framework than SBF. | ['Francesco Archetti', 'Andrea Ponti', 'Antonio Candelieri'] | 2022-10-12 | null | null | null | null | ['marketing'] | ['miscellaneous'] | [ 6.81639463e-02 -4.38657939e-01 -6.92479312e-01 -2.26213470e-01
-9.36111987e-01 -4.03032303e-01 5.04490197e-01 1.26874939e-01
-6.62419081e-01 1.08509612e+00 2.03402176e-01 -8.03190529e-01
-1.00824773e+00 -5.12905896e-01 -4.71181005e-01 -8.59034538e-01
-3.95258293e-02 7.90818810e-01 -1.04486659e-01 1.62539899e-01
4.71524209e-01 7.93853462e-01 -1.34912419e+00 -1.86014697e-01
5.22292435e-01 1.33893466e+00 5.12260318e-01 4.37965423e-01
-1.56757757e-01 4.12006795e-01 -3.76237929e-01 -3.01047653e-01
5.45950651e-01 -1.02225907e-01 -1.08702552e+00 1.31845862e-01
-3.51401567e-01 -4.08908963e-01 1.18127696e-01 9.59726155e-01
4.85615134e-01 4.10648614e-01 5.64648986e-01 -1.11735308e+00
-7.14654848e-02 8.47359419e-01 -4.61131483e-01 4.60159987e-01
2.12869003e-01 -1.30070224e-01 1.17227650e+00 -5.28733253e-01
2.50068873e-01 1.34393156e+00 1.71242982e-01 3.59957755e-01
-1.28081155e+00 -3.33557785e-01 2.45824680e-01 4.19936925e-01
-1.33934116e+00 -3.82998377e-01 5.17943799e-01 -5.41205287e-01
6.63874865e-01 5.87469518e-01 8.66061509e-01 7.02778876e-01
-8.46531317e-02 9.04334486e-01 1.41510451e+00 -9.44730937e-01
5.57468355e-01 4.20824111e-01 3.24885547e-01 6.23395517e-02
4.08220708e-01 3.13799173e-01 -5.30885220e-01 -5.53438187e-01
5.21871448e-01 7.25063160e-02 -2.34942913e-01 -4.70437169e-01
-8.66048574e-01 1.09825373e+00 2.15926282e-02 2.44261086e-01
-8.16697776e-01 4.35669601e-01 3.79907750e-02 2.18630984e-01
6.20401621e-01 4.04488057e-01 -3.64217579e-01 -6.09805048e-01
-1.13627434e+00 2.92139530e-01 8.94413710e-01 6.13703966e-01
5.77438593e-01 9.06647742e-03 -6.10162973e-01 7.21412063e-01
8.38061512e-01 4.51331407e-01 3.32505196e-01 -9.02511239e-01
1.48714930e-01 -1.86482951e-01 6.87133193e-01 -8.92184019e-01
-2.32855290e-01 -6.93731546e-01 -5.27673721e-01 -2.16865718e-01
6.10882640e-01 -4.82131653e-02 -4.38501030e-01 1.48794281e+00
2.45852679e-01 1.25280991e-01 -2.64410287e-01 1.05732489e+00
3.63298208e-02 7.49647677e-01 -3.10699105e-01 -9.54656839e-01
1.16139448e+00 -8.01391602e-01 -9.66583133e-01 -3.79004665e-02
2.87309170e-01 -8.57713044e-01 7.47918546e-01 8.84316385e-01
-9.10620570e-01 5.09317853e-02 -7.38102376e-01 7.35332131e-01
-2.58743286e-01 -1.65064439e-01 9.54291165e-01 1.17680597e+00
-1.05862617e+00 5.72088122e-01 -4.50168371e-01 -3.08979064e-01
3.10221910e-01 3.83160472e-01 2.72415847e-01 -2.47079074e-01
-1.37579322e+00 9.59283590e-01 2.70231128e-01 2.23978594e-01
-9.06858027e-01 -6.41771793e-01 -3.00529182e-01 2.67142385e-01
6.93763137e-01 -4.87950385e-01 1.41256690e+00 -8.92194271e-01
-1.94871724e+00 4.01263267e-01 1.14882708e-01 -6.42159939e-01
6.50475740e-01 -3.26862693e-01 -1.21446010e-02 -1.78132460e-01
-2.37277344e-01 3.85027081e-02 1.07672298e+00 -1.01163030e+00
-7.23646104e-01 -3.78576756e-01 1.07831091e-01 1.86185285e-01
-4.42445964e-01 2.49721467e-01 -1.48124635e-01 -5.49116671e-01
-2.09685653e-01 -8.92027438e-01 -4.95843709e-01 -6.79261565e-01
-4.17469889e-01 -2.64780045e-01 3.26430589e-01 -5.95351398e-01
1.89736962e+00 -1.79851270e+00 1.89774036e-01 7.65594482e-01
-3.41684431e-01 1.19267488e-02 1.56998411e-01 3.87962699e-01
1.12285856e-02 3.24547321e-01 1.11761473e-01 -2.27302000e-01
2.77395487e-01 3.46978009e-01 -1.43343687e-01 5.77868164e-01
-2.68918276e-01 3.00805032e-01 -9.26582158e-01 -2.43000999e-01
2.83018798e-01 -5.25022806e-05 -4.20295894e-01 3.41116823e-02
-5.66915125e-02 3.04551184e-01 -5.73844790e-01 5.08505046e-01
4.63253707e-01 -3.00554484e-01 1.65517703e-01 8.78635049e-02
-1.99368969e-01 -1.20452076e-01 -1.49724364e+00 1.46221602e+00
-7.30021179e-01 5.46715558e-01 2.15051800e-01 -1.28099275e+00
7.70976067e-01 2.69064456e-01 9.54977155e-01 -4.88121361e-01
1.97564557e-01 2.81287938e-01 1.01558469e-01 -4.94844645e-01
5.12694657e-01 -1.83719635e-01 -1.96100120e-02 6.21787310e-01
-1.03657827e-01 6.07359037e-02 3.85902077e-01 7.90555105e-02
8.04762721e-01 -8.91499966e-02 5.56773126e-01 -6.51768565e-01
3.91559154e-01 -3.91772777e-01 3.70429516e-01 1.25406134e+00
-3.11506361e-01 1.08705917e-02 2.92999566e-01 -2.06817791e-01
-6.81064487e-01 -5.75684190e-01 -4.35429126e-01 1.22765124e+00
-1.35074100e-02 -3.71501036e-02 -6.69114351e-01 -3.90516579e-01
3.60561550e-01 7.64577746e-01 -4.23000634e-01 2.15696022e-01
2.33280472e-02 -1.14722288e+00 -2.58333385e-01 -3.81965972e-02
2.57867962e-01 -5.53600013e-01 -9.76852357e-01 4.81295913e-01
-2.61642814e-01 -7.87325621e-01 -2.98819423e-01 3.39180678e-01
-7.87691593e-01 -7.28416622e-01 -1.00747049e+00 4.36307549e-01
-9.30645224e-03 2.31098935e-01 1.05704319e+00 -3.75928521e-01
-2.72600085e-01 9.07744706e-01 -4.94054258e-01 -7.15102911e-01
-1.46906033e-01 -8.90225098e-02 1.26416728e-01 5.40247440e-01
2.20828474e-01 -9.62059647e-02 -5.21839857e-01 3.22316796e-01
-8.03203940e-01 -3.05278838e-01 6.04493439e-01 1.05389094e+00
5.45349419e-01 1.31564915e-01 6.76714301e-01 -1.10621202e+00
8.98549795e-01 -6.74253523e-01 -1.04790783e+00 3.83881778e-01
-1.12618434e+00 -9.05380547e-02 5.22822142e-02 -6.76266328e-02
-1.16515124e+00 -2.71461487e-01 1.84642226e-01 -4.95509058e-01
9.31769796e-03 9.90894794e-01 3.47492211e-02 -4.18970883e-02
4.27039027e-01 -7.55160451e-02 4.31634998e-03 -5.97373068e-01
2.33398542e-01 9.74934578e-01 -1.99086681e-01 -8.15163791e-01
2.45338872e-01 2.92366385e-01 2.61248618e-01 -6.45339608e-01
-9.61738110e-01 -9.19134617e-01 -3.76543522e-01 -5.45901239e-01
4.42765296e-01 -4.31609511e-01 -7.52353430e-01 8.52474123e-02
-1.00600231e+00 -4.04139578e-01 -2.58019835e-01 8.42486501e-01
-8.65165353e-01 2.08917066e-01 -1.88994721e-01 -1.54872417e+00
-1.07124969e-01 -1.29000401e+00 7.41651297e-01 1.87158599e-01
-3.49035300e-02 -1.03728127e+00 -3.24619524e-02 5.17012835e-01
6.43575251e-01 -9.54408050e-02 7.45785654e-01 -5.96497595e-01
-6.43616021e-01 -9.00770798e-02 -1.01380177e-01 3.44471484e-01
-1.04731105e-01 -4.51857969e-02 -8.46786261e-01 -6.77491128e-01
1.69170238e-02 1.76119488e-02 6.13811314e-01 8.29409361e-01
1.23110485e+00 -4.84384418e-01 -1.03224374e-01 2.07878366e-01
1.69624484e+00 4.33131427e-01 2.42485508e-01 5.68142831e-01
-1.10697836e-01 6.71834111e-01 1.16480899e+00 1.21067011e+00
-1.69091433e-01 1.12152767e+00 7.16958344e-01 2.86310136e-01
3.41432273e-01 2.51868278e-01 1.88873589e-01 7.59193957e-01
-9.78925973e-02 -2.97645420e-01 -9.73274767e-01 4.24092114e-01
-2.32067108e+00 -7.39738524e-01 7.23009780e-02 2.65382338e+00
4.75581050e-01 -2.95324181e-03 3.45929831e-01 2.19758779e-01
6.48481607e-01 -4.59319986e-02 -4.61138785e-01 -7.20767498e-01
1.55480996e-01 9.40253884e-02 8.95330846e-01 5.27425170e-01
-1.07182682e+00 5.03195584e-01 6.60213232e+00 1.25289989e+00
-8.84514630e-01 4.15765703e-01 9.26279068e-01 -2.61641353e-01
-1.83541536e-01 5.41996434e-02 -7.05903828e-01 7.29264915e-01
1.18876982e+00 -2.39132389e-01 8.96847785e-01 7.55217910e-01
6.12941325e-01 -5.45833290e-01 -8.21803868e-01 1.27377272e+00
-2.92748004e-01 -1.28747642e+00 -2.81702131e-01 2.01932192e-01
6.91600740e-01 -1.98003918e-01 8.89431909e-02 1.36489064e-01
3.12288016e-01 -9.78269875e-01 8.62774372e-01 8.29813659e-01
5.00700712e-01 -7.99204528e-01 1.09073734e+00 4.50773090e-01
-4.63629037e-01 -4.40180361e-01 -2.15653673e-01 -2.12637857e-01
2.14944646e-01 9.08908963e-01 -5.81013978e-01 7.97866702e-01
1.03377593e+00 5.20402074e-01 2.80917552e-03 1.40833080e+00
1.69944212e-01 7.75020063e-01 -3.76237869e-01 -4.33708251e-01
5.13758183e-01 -4.10029501e-01 6.85002983e-01 1.25569904e+00
4.58391547e-01 -8.62895772e-02 2.43307710e-01 4.68756944e-01
4.32572275e-01 4.11215812e-01 -3.44411463e-01 -2.75089759e-02
6.75173700e-01 1.11336946e+00 -6.12650514e-01 -1.50027514e-01
-3.78397882e-01 2.23647177e-01 -1.12448350e-01 5.74996054e-01
-5.56450546e-01 -1.07616276e-01 5.98410130e-01 -6.94701448e-02
2.31056437e-01 1.27465785e-01 -3.84363681e-02 -7.07924664e-01
-3.32767457e-01 -6.22846663e-01 5.52650213e-01 -5.70187271e-01
-1.19280171e+00 1.60958171e-01 5.12680292e-01 -9.21892166e-01
-3.11308175e-01 -7.34169543e-01 1.79390177e-01 1.00508749e+00
-1.80817235e+00 -5.69774568e-01 8.93419534e-02 5.28347850e-01
4.90776122e-01 -2.76560694e-01 6.80524707e-01 3.92132401e-01
-7.23922074e-01 2.04628870e-01 8.99635553e-01 -6.26765251e-01
4.41387266e-01 -1.41879582e+00 -9.41743255e-01 7.20387638e-01
1.07413389e-01 4.11541760e-01 9.20203149e-01 -2.01348811e-01
-1.37141383e+00 -7.10001767e-01 3.42768341e-01 1.59309030e-01
7.66696930e-01 -5.67479469e-02 -4.13757950e-01 3.24291855e-01
1.78975776e-01 -3.54294121e-01 9.23902214e-01 3.84964585e-01
2.33688608e-01 -4.94888902e-01 -1.13196731e+00 2.47065276e-01
5.28413892e-01 -1.32774666e-01 -1.23964320e-03 7.08441198e-01
2.09838778e-01 -2.10407525e-01 -1.03716528e+00 2.64368147e-01
5.76606691e-01 -9.75901186e-01 7.53806531e-01 -5.84753275e-01
-1.66296512e-01 -4.37827548e-03 -5.36331177e-01 -1.37176251e+00
-3.97742450e-01 -1.25305843e+00 -3.35433453e-01 1.02391267e+00
1.81181371e-01 -7.86136389e-01 7.52694309e-01 5.00172913e-01
2.91421145e-01 -1.08963776e+00 -1.35491621e+00 -9.17960286e-01
-1.44069493e-01 -6.41476929e-01 5.41806757e-01 5.92238665e-01
-1.81041569e-01 3.92430536e-02 -6.23783529e-01 -1.34891927e-01
5.77660859e-01 2.44673222e-01 6.29064977e-01 -1.21999156e+00
-6.57857478e-01 -7.17238605e-01 -1.11248367e-01 -1.12074161e+00
4.54215370e-02 -5.28928161e-01 1.15373805e-01 -1.20922923e+00
3.80634725e-01 -7.22372413e-01 -7.57868886e-01 7.38143325e-02
2.03383029e-01 -1.65839523e-01 3.95173132e-01 3.11720997e-01
-4.86723572e-01 3.91121984e-01 1.00367856e+00 -1.36866178e-02
1.06150255e-01 6.43544197e-01 -6.03880286e-01 4.65316653e-01
5.59720695e-01 -5.43470860e-01 -4.25392538e-01 -3.82965393e-02
5.04470050e-01 5.59384048e-01 6.63021877e-02 -3.55699152e-01
1.10288493e-01 -6.02910995e-01 -2.46777177e-01 -3.77332002e-01
4.23068374e-01 -1.14756989e+00 2.40367562e-01 3.75663370e-01
-3.51776183e-01 -1.41230270e-01 1.33613432e-02 8.94817889e-01
-2.23893642e-01 -9.38479900e-01 7.08743691e-01 -1.59172863e-01
-7.38404691e-01 1.92932293e-01 -5.59723914e-01 -2.76100129e-01
9.94189680e-01 -1.57355204e-01 2.03339830e-02 -7.75823295e-01
-1.08444369e+00 2.86560893e-01 -2.01110616e-01 4.07356918e-02
3.72740775e-01 -1.10954106e+00 -5.96669555e-01 -3.16011459e-01
-2.22336471e-01 -4.33451891e-01 2.50718832e-01 1.11007929e+00
-3.07002604e-01 9.95659232e-01 8.37012231e-02 -7.59268582e-01
-1.10814714e+00 4.47960347e-01 2.29308203e-01 -7.79049456e-01
2.72081345e-01 7.30361879e-01 -4.77314740e-01 2.22850479e-02
3.81932467e-01 8.19149688e-02 -2.43585140e-01 4.41420346e-01
2.36008003e-01 7.19677627e-01 2.48865336e-01 -2.48981580e-01
-2.39830673e-01 9.84477699e-02 1.73713356e-01 -5.29346049e-01
1.34590578e+00 -5.19525051e-01 -1.82074547e-01 7.31807709e-01
6.97282255e-01 -2.49039024e-01 -9.84585762e-01 -5.20595074e-01
5.17040253e-01 -9.69348967e-01 6.06334865e-01 -8.60550702e-01
-8.98576677e-01 7.76200652e-01 7.91052878e-01 8.58339429e-01
1.17891014e+00 -3.47609282e-01 -3.22189815e-02 2.73177952e-01
7.92406082e-01 -1.35122669e+00 2.29906999e-02 2.57068694e-01
8.31773400e-01 -1.13580167e+00 2.11203232e-01 -4.95751165e-02
-2.75305271e-01 1.13248444e+00 3.69438082e-02 2.53054082e-01
1.14043033e+00 -3.10639329e-02 -2.65953869e-01 2.73797691e-01
-7.17685461e-01 -2.59570360e-01 1.02238558e-01 2.82261401e-01
2.63774037e-01 3.73478293e-01 -7.17027426e-01 5.99578440e-01
1.99513108e-01 3.63629172e-03 5.31979501e-01 9.06593740e-01
-2.63954610e-01 -1.07879889e+00 -6.42452121e-01 8.25931251e-01
-7.68669963e-01 4.84401546e-02 1.12403177e-01 7.46218443e-01
-2.63067722e-01 1.17439699e+00 -1.16482623e-01 1.17276318e-01
8.78984481e-02 -1.03607491e-01 5.44068933e-01 -5.72379053e-01
-4.46576655e-01 4.89897162e-01 8.12949389e-02 -6.75641119e-01
-8.89276624e-01 -1.07444906e+00 -2.90824354e-01 -1.29448041e-01
-8.51700902e-01 5.45261741e-01 1.05441463e+00 9.49569106e-01
1.27581462e-01 6.58235967e-01 9.11155581e-01 -8.59771609e-01
-1.15485656e+00 -1.05315888e+00 -1.09528089e+00 2.20595434e-01
1.44159496e-01 -9.49230850e-01 -2.22100735e-01 -4.80150372e-01] | [4.55596923828125, 3.2507448196411133] |
a9e411fa-17d8-4b28-8dca-51de0d05f542 | estimating-the-amenibility-of-new-domains-for | null | null | https://aclanthology.org/W16-0804 | https://aclanthology.org/W16-0804.pdf | Estimating the amenibility of new domains for deception detection | null | ['Eileen Fitzpatrick', 'Joan Bachenko'] | 2016-06-01 | null | null | null | ws-2016-6 | ['deception-detection'] | ['miscellaneous'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.357299327850342, 3.717125654220581] |
0eb8fb1d-7355-4ed8-9d9c-d14b4ee1fed6 | exploiting-simulated-user-feedback-for | 2304.13874 | null | https://arxiv.org/abs/2304.13874v3 | https://arxiv.org/pdf/2304.13874v3.pdf | Exploiting Simulated User Feedback for Conversational Search: Ranking, Rewriting, and Beyond | This research aims to explore various methods for assessing user feedback in mixed-initiative conversational search (CS) systems. While CS systems enjoy profuse advancements across multiple aspects, recent research fails to successfully incorporate feedback from the users. One of the main reasons for that is the lack of system-user conversational interaction data. To this end, we propose a user simulator-based framework for multi-turn interactions with a variety of mixed-initiative CS systems. Specifically, we develop a user simulator, dubbed ConvSim, that, once initialized with an information need description, is capable of providing feedback to a system's responses, as well as answering potential clarifying questions. Our experiments on a wide variety of state-of-the-art passage retrieval and neural re-ranking models show that effective utilization of user feedback can lead to 16% retrieval performance increase in terms of nDCG@3. Moreover, we observe consistent improvements as the number of feedback rounds increases (35% relative improvement in terms of nDCG@3 after three rounds). This points to a research gap in the development of specific feedback processing modules and opens a potential for significant advancements in CS. To support further research in the topic, we release over 30,000 transcripts of system-simulator interactions based on well-established CS datasets. | ['Fabio Crestani', 'Jeffrey Dalton', 'Mohammad Aliannejadi', 'Ivan Sekulić', 'Paul Owoicho'] | 2023-04-26 | null | null | null | null | ['passage-retrieval', 'conversational-search'] | ['natural-language-processing', 'natural-language-processing'] | [ 1.11223586e-01 -9.68805626e-02 -1.05835781e-01 -3.60188931e-01
-1.42877126e+00 -8.61473739e-01 8.16579282e-01 3.58690917e-01
-6.01484537e-01 5.95539212e-01 6.46264672e-01 -6.90959632e-01
-1.24747343e-01 -1.60805464e-01 -2.46603459e-01 -8.27276036e-02
1.40466109e-01 5.45980155e-01 3.70442837e-01 -9.40524936e-01
8.38511407e-01 1.13007754e-01 -1.52160823e+00 6.65821731e-01
1.05610538e+00 4.83969599e-01 4.96046066e-01 1.21301365e+00
-2.50752151e-01 8.11802745e-01 -8.68620455e-01 -3.38224351e-01
-2.94170171e-01 -3.91488880e-01 -1.43706775e+00 -5.37627399e-01
4.32378381e-01 -5.20633638e-01 -2.69272327e-01 5.48711419e-01
8.99082482e-01 4.70554560e-01 4.51806158e-01 -1.07912886e+00
-6.04546368e-01 7.19766259e-01 8.28810930e-02 5.77861786e-01
1.19017434e+00 9.43664759e-02 1.38429403e+00 -8.08618784e-01
6.73235536e-01 1.18277168e+00 2.11262479e-01 5.22095561e-01
-9.12402153e-01 -3.50858003e-01 -1.62106212e-02 3.24824750e-01
-9.79268014e-01 -4.85201150e-01 4.30867642e-01 -1.84563681e-01
1.32567620e+00 7.99147427e-01 2.81433254e-01 9.97610092e-01
-1.16695590e-01 1.32105958e+00 8.02060127e-01 -7.18383074e-01
-8.51410329e-02 4.23918188e-01 7.45406389e-01 2.67147422e-01
-3.06352049e-01 -2.17837110e-01 -6.07837677e-01 -3.98419827e-01
1.91545904e-01 -2.76895732e-01 -6.42283976e-01 1.59202829e-01
-7.44106770e-01 6.35023773e-01 2.81902581e-01 5.03722548e-01
-2.05577850e-01 -2.70507067e-01 4.73848403e-01 6.85772657e-01
4.78432834e-01 9.56691086e-01 -4.95034188e-01 -8.82542789e-01
-7.35323966e-01 6.03714705e-01 1.30321598e+00 8.68308246e-01
3.70402724e-01 -7.31823921e-01 -6.92342877e-01 1.25036144e+00
1.41532823e-01 2.99424857e-01 5.39865673e-01 -1.00955749e+00
5.99095881e-01 6.85229957e-01 5.13716698e-01 -9.77843225e-01
-2.28431627e-01 -3.90739352e-01 -1.33515865e-01 -5.49364448e-01
7.41239563e-02 4.46472829e-03 -3.56446236e-01 1.60828710e+00
-1.62567824e-01 -2.95727372e-01 5.40870391e-02 7.33542621e-01
1.06905413e+00 5.45365989e-01 -6.81167096e-02 -3.17223310e-01
1.22191942e+00 -1.20056105e+00 -7.26883173e-01 -1.49549380e-01
9.59469914e-01 -1.01572943e+00 1.56063139e+00 2.36354947e-01
-1.39301097e+00 -4.39496368e-01 -7.11728454e-01 -1.10057853e-01
-2.32802153e-01 -5.05401902e-02 4.48926181e-01 6.08838677e-01
-1.53341341e+00 3.93118411e-01 -4.74724680e-01 -6.03758514e-01
-4.79512304e-01 1.87988400e-01 2.38624290e-01 -1.76574841e-01
-1.40341187e+00 1.05328512e+00 -3.19111794e-01 -1.25925168e-01
-4.18351442e-01 -7.12241590e-01 -4.51540023e-01 4.06018019e-01
3.20561558e-01 -6.82958007e-01 2.24829412e+00 -3.74243557e-01
-1.53136122e+00 4.79127586e-01 -3.73679996e-01 1.84337271e-03
3.76228094e-01 -5.81245482e-01 -3.84485751e-01 8.84901434e-02
6.83581829e-02 4.23690856e-01 4.14758287e-02 -1.11025560e+00
-7.49086738e-01 -1.24899827e-01 5.66538990e-01 8.75613570e-01
-6.11089766e-01 4.14497077e-01 -8.62988114e-01 -1.18416503e-01
-3.45584780e-01 -9.76917922e-01 -1.61130548e-01 -7.54190803e-01
-1.61652029e-01 -6.17460549e-01 4.22480345e-01 -6.71975613e-01
1.88977325e+00 -1.67474329e+00 1.09076267e-02 -2.14550849e-02
7.04681277e-02 5.08411050e-01 -3.67891520e-01 1.12681055e+00
2.98168212e-01 4.33608204e-01 2.42580816e-01 -3.88551980e-01
-3.31405774e-02 -3.69455278e-01 -1.37881830e-01 -3.63221347e-01
-2.56658524e-01 9.64564264e-01 -1.23767984e+00 -2.90747851e-01
-6.80683404e-02 1.82994142e-01 -7.56640494e-01 5.85650206e-01
-2.16557011e-01 2.46843696e-01 -5.77900469e-01 3.53799462e-01
3.32259446e-01 -5.22933483e-01 1.19574657e-02 2.65102416e-01
-1.65648490e-01 1.01786697e+00 -7.39115655e-01 1.80791223e+00
-6.77569747e-01 7.56808043e-01 5.40090427e-02 -4.08048153e-01
4.14687783e-01 5.38233161e-01 2.82104135e-01 -1.07981217e+00
-5.88146923e-03 3.80068570e-02 -7.12353662e-02 -6.04507565e-01
1.26592028e+00 6.07812464e-01 1.79774128e-02 1.11937439e+00
-2.60819882e-01 -8.17492232e-02 4.49877232e-01 8.56122494e-01
1.43772602e+00 -5.95584035e-01 -1.80735264e-03 -1.12322986e-01
6.41821265e-01 6.08466798e-03 -2.49064684e-01 1.21633470e+00
-1.72736168e-01 3.38214040e-01 1.05443224e-01 1.44007951e-01
-5.41208982e-01 -6.03130639e-01 2.05970973e-01 1.68265343e+00
9.21631157e-02 -6.65608883e-01 -9.54552114e-01 -5.19320488e-01
-2.29437232e-01 9.33933794e-01 -2.05000743e-01 -1.60104364e-01
-3.92479748e-01 -2.87700415e-01 5.23785353e-01 4.15296346e-01
3.88331503e-01 -1.31410515e+00 -3.86180222e-01 2.42963582e-01
-9.07292306e-01 -8.88289511e-01 -7.90235817e-01 -2.97489762e-01
-7.82721579e-01 -1.08284831e+00 -7.93808699e-01 -6.44606411e-01
2.98592836e-01 1.07775426e+00 1.48066580e+00 6.65457129e-01
-1.14868239e-01 1.01416290e+00 -6.72509909e-01 -1.42041028e-01
-5.47589362e-01 3.80293816e-01 -7.79406577e-02 -8.00153792e-01
4.93950188e-01 -1.89320743e-01 -9.66955900e-01 5.40706635e-01
-8.01938474e-01 -2.23793551e-01 4.22509462e-01 6.37781382e-01
-3.12484324e-01 -3.56422663e-01 6.00419939e-01 -8.67714763e-01
1.97586513e+00 -5.12449443e-01 -5.65072782e-02 6.61528945e-01
-9.59931195e-01 3.70265096e-02 1.76162168e-01 -2.57638246e-01
-1.34853017e+00 -7.35831439e-01 -1.97139248e-01 3.22800994e-01
1.42446076e-02 8.73867750e-01 5.14139414e-01 1.09077126e-01
1.04246974e+00 2.95221686e-01 -2.41425246e-01 -3.81071508e-01
4.80233997e-01 1.31377029e+00 2.70715415e-01 -7.21692204e-01
1.84321404e-01 -2.89597422e-01 -1.11008072e+00 -8.68990123e-01
-5.29014647e-01 -1.17453313e+00 -1.38431117e-01 -4.70989972e-01
3.33129376e-01 -6.79136217e-01 -1.05371761e+00 2.21404746e-01
-1.17804050e+00 -4.84483421e-01 3.83174956e-01 2.53024727e-01
-3.05503458e-01 6.02547646e-01 -9.83576238e-01 -1.14044690e+00
-8.20050001e-01 -1.27964938e+00 8.24480236e-01 4.58976358e-01
-6.83861554e-01 -8.83234680e-01 3.92718375e-01 9.36043143e-01
8.06511343e-01 -8.09103251e-01 8.96764219e-01 -7.36969292e-01
-4.94983941e-01 -3.03308994e-01 -1.22453779e-01 2.69610703e-01
6.56376854e-02 2.22984441e-02 -8.07237864e-01 -5.94600797e-01
-1.86004221e-01 -6.23247504e-01 4.50162202e-01 1.88985005e-01
8.93347204e-01 -2.62361676e-01 -3.72763425e-01 -3.88836771e-01
8.90081644e-01 4.48245853e-01 4.96215940e-01 2.86498368e-01
2.72894185e-02 5.26854694e-01 7.30700791e-01 3.39989334e-01
6.81895494e-01 7.57962048e-01 -2.10404713e-02 1.67882010e-01
1.04789346e-01 -2.71668464e-01 3.98252398e-01 1.08049762e+00
4.53565083e-02 -7.00515270e-01 -7.96211839e-01 5.42557776e-01
-1.70780706e+00 -8.55263889e-01 -6.84943348e-02 2.24003792e+00
9.08855557e-01 1.26464337e-01 1.51776254e-01 -1.47989348e-01
3.45409185e-01 -2.49158628e-02 -3.50407392e-01 -6.38501227e-01
4.55479920e-01 1.26936913e-01 -3.77943777e-02 7.96132803e-01
-3.73971820e-01 8.92909110e-01 6.90553570e+00 4.35063839e-01
-8.39148998e-01 -1.68314546e-01 5.18858492e-01 1.06421977e-01
-5.25811434e-01 1.67086604e-04 -7.61270404e-01 3.94591510e-01
1.19395065e+00 -5.60226381e-01 5.97141027e-01 7.79276967e-01
2.93909490e-01 -3.28384161e-01 -1.15781701e+00 7.13306606e-01
1.82831094e-01 -1.24981225e+00 -5.72022907e-02 -2.39026979e-01
4.25157845e-01 2.09323660e-01 -2.20291018e-02 8.91340911e-01
6.84134901e-01 -5.65347552e-01 1.40564278e-01 3.71926844e-01
4.86491203e-01 -5.47708631e-01 6.97619438e-01 7.58696258e-01
-7.88786054e-01 5.91664463e-02 -1.13997184e-01 -2.76281953e-01
2.25030705e-01 -4.91495132e-02 -1.44146168e+00 5.52804828e-01
7.08821893e-01 2.15433747e-01 -6.72229111e-01 1.10223687e+00
-4.52452190e-02 6.95209682e-01 -2.24076912e-01 -6.69542670e-01
1.95316732e-01 2.31462959e-02 3.50738585e-01 1.50057745e+00
7.16647059e-02 4.85166728e-01 1.53629184e-01 4.37770665e-01
-1.77178428e-01 4.68565285e-01 -3.92289877e-01 -2.39434808e-01
8.78785312e-01 1.10340369e+00 -2.88735181e-01 -4.82916921e-01
-3.36584866e-01 1.16630793e+00 4.36333537e-01 5.92313945e-01
-3.11038077e-01 -3.42808992e-01 4.74880069e-01 -6.64218888e-02
-2.84721613e-01 -1.61124632e-01 2.61928812e-02 -9.34477270e-01
1.81345224e-01 -1.21001875e+00 4.75026011e-01 -9.57014084e-01
-1.02443397e+00 6.85493827e-01 1.06910437e-01 -8.70628655e-01
-7.68213630e-01 -1.03033893e-01 -7.58184135e-01 1.07610607e+00
-1.44227219e+00 -5.67390084e-01 -4.18670207e-01 2.84903437e-01
1.04733348e+00 5.46041355e-02 9.93775129e-01 4.50944275e-01
-3.76602262e-01 9.07389343e-01 1.06891893e-01 -2.36754403e-01
1.08594120e+00 -1.19703519e+00 5.13215899e-01 3.96951914e-01
1.20902367e-01 1.35098636e+00 8.55295122e-01 -5.91596246e-01
-1.36946952e+00 -1.85071841e-01 1.20077908e+00 -8.14528883e-01
4.82051253e-01 -1.34915233e-01 -9.21360254e-01 4.09253120e-01
6.45453691e-01 -8.03440869e-01 8.42920542e-01 7.42160678e-01
-7.45178340e-03 3.48484397e-01 -8.10825646e-01 9.96342838e-01
8.59981060e-01 -9.16316926e-01 -5.39355159e-01 4.72979933e-01
9.89879191e-01 -5.46932042e-01 -7.77180910e-01 2.77379185e-01
5.40160120e-01 -8.26140165e-01 9.31512892e-01 -7.71129191e-01
3.45596224e-01 2.35462338e-01 6.21054582e-02 -1.54371977e+00
-3.06363374e-01 -8.12285006e-01 -3.72281410e-02 1.18509936e+00
6.93354905e-01 -3.04354906e-01 6.51962519e-01 1.13867259e+00
-2.82040477e-01 -7.35244393e-01 -3.84282559e-01 -2.41902858e-01
-2.35398524e-02 -4.15429175e-01 2.90507674e-01 7.73998380e-01
7.11862803e-01 7.68569708e-01 -9.47597343e-03 -1.80075407e-01
-1.72734395e-01 -1.67687014e-01 1.02911580e+00 -1.07366550e+00
-1.95287630e-01 -6.44629002e-01 3.65223348e-01 -1.77781332e+00
-1.51236311e-01 -7.05569625e-01 1.10093310e-01 -1.68534482e+00
5.27166247e-01 -4.24060255e-01 -4.30951446e-01 9.75749269e-02
-6.09203160e-01 -3.57451737e-01 9.94810537e-02 3.30113292e-01
-9.70924079e-01 4.27879810e-01 1.13551366e+00 -1.55140134e-02
-6.06888175e-01 3.54519486e-01 -1.08514583e+00 1.66470379e-01
4.89121675e-01 -1.23018309e-01 -7.69711971e-01 -4.54002529e-01
3.44941467e-01 5.83078563e-01 -2.57123113e-01 -7.02280104e-01
7.85863996e-01 -5.17107546e-02 -3.98204476e-01 -8.49363863e-01
2.66014665e-01 -3.07767123e-01 -4.11494702e-01 1.01872586e-01
-1.08863544e+00 4.74889249e-01 2.42154747e-01 4.25506949e-01
-3.04656476e-01 -2.69084454e-01 2.03725626e-03 -2.06758305e-01
-5.36076248e-01 -1.24608792e-01 -6.55313611e-01 1.06943771e-01
3.01089108e-01 6.29335409e-04 -5.70278287e-01 -1.10623562e+00
-1.71810105e-01 7.31093705e-01 1.92726657e-01 8.31858337e-01
4.64853913e-01 -8.10000420e-01 -6.63432181e-01 -2.43097514e-01
5.08025050e-01 -4.76641327e-01 3.52839082e-01 5.65884471e-01
-1.13161407e-01 1.14612055e+00 1.89475209e-01 -6.26130819e-01
-1.60957730e+00 -6.13609403e-02 5.30155422e-03 -6.45700574e-01
-1.88727647e-01 9.82077479e-01 -2.49951959e-01 -7.74858892e-01
5.66312611e-01 -2.67496109e-01 -4.81287837e-01 4.31295410e-02
8.21417093e-01 3.77196133e-01 5.87644279e-01 -9.33698043e-02
-3.16163301e-02 -1.89160943e-01 -7.00540543e-01 -4.96017635e-01
9.52276945e-01 -3.86741310e-01 1.41231969e-01 2.92008251e-01
1.29416943e+00 -1.42792342e-02 -5.47022104e-01 -5.70895672e-01
2.63903052e-01 -4.53128666e-01 1.18969977e-02 -1.37847102e+00
-4.24298435e-01 6.32699668e-01 5.22053301e-01 4.04294401e-01
8.65447402e-01 -1.60077825e-01 8.89312327e-01 1.05605412e+00
3.37284803e-01 -1.23578918e+00 4.14761662e-01 7.95434356e-01
1.00476646e+00 -1.38686216e+00 -3.34390253e-01 -2.73143388e-02
-7.61508703e-01 7.50872135e-01 7.08005071e-01 4.83737916e-01
2.82501608e-01 -2.05184847e-01 4.20160383e-01 -4.10914570e-01
-1.17592299e+00 1.45857977e-02 3.62938851e-01 5.63912913e-02
1.13060641e+00 1.50551219e-02 -7.00827360e-01 3.38885874e-01
-3.27020645e-01 8.12188685e-02 4.88345891e-01 1.11215532e+00
-5.26940882e-01 -1.28433084e+00 2.92976461e-02 4.68915999e-01
-3.34919125e-01 -4.13626641e-01 -7.93931305e-01 4.83726680e-01
-1.10769451e+00 1.56222606e+00 -2.81061053e-01 -6.10492051e-01
6.32811069e-01 2.18427837e-01 4.33930792e-02 -7.16248572e-01
-1.01898551e+00 -2.86981553e-01 5.23183525e-01 -7.21382141e-01
-1.41913801e-01 -5.88373959e-01 -9.15540993e-01 -2.31771961e-01
-6.93139613e-01 8.22736382e-01 8.64136994e-01 8.36813569e-01
5.12566805e-01 4.22563702e-01 6.80402160e-01 -4.93451118e-01
-8.39404762e-01 -1.48282254e+00 6.44076541e-02 5.63709259e-01
1.09122753e-01 -4.36842620e-01 -3.97024274e-01 -4.29247022e-01] | [12.12612247467041, 7.818840980529785] |
d9cc975f-f034-44e7-91ff-b0443e51b277 | knowledge-adaptation-teaching-to-adapt | 1702.02052 | null | http://arxiv.org/abs/1702.02052v1 | http://arxiv.org/pdf/1702.02052v1.pdf | Knowledge Adaptation: Teaching to Adapt | Domain adaptation is crucial in many real-world applications where the
distribution of the training data differs from the distribution of the test
data. Previous Deep Learning-based approaches to domain adaptation need to be
trained jointly on source and target domain data and are therefore unappealing
in scenarios where models need to be adapted to a large number of domains or
where a domain is evolving, e.g. spam detection where attackers continuously
change their tactics.
To fill this gap, we propose Knowledge Adaptation, an extension of Knowledge
Distillation (Bucilua et al., 2006; Hinton et al., 2015) to the domain
adaptation scenario. We show how a student model achieves state-of-the-art
results on unsupervised domain adaptation from multiple sources on a standard
sentiment analysis benchmark by taking into account the domain-specific
expertise of multiple teachers and the similarities between their domains.
When learning from a single teacher, using domain similarity to gauge
trustworthiness is inadequate. To this end, we propose a simple metric that
correlates well with the teacher's accuracy in the target domain. We
demonstrate that incorporating high-confidence examples selected by this metric
enables the student model to achieve state-of-the-art performance in the
single-source scenario. | ['John G. Breslin', 'Sebastian Ruder', 'Parsa Ghaffari'] | 2017-02-07 | null | null | null | null | ['spam-detection'] | ['natural-language-processing'] | [ 4.22852626e-03 1.03550762e-01 -2.39641294e-01 -5.19533992e-01
-7.31925070e-01 -1.14213145e+00 8.14538598e-01 5.10103166e-01
-6.43621445e-01 8.59859765e-01 -5.84175214e-02 -4.20452327e-01
-1.82844535e-01 -7.17038453e-01 -8.08257878e-01 -4.61746275e-01
4.50680226e-01 8.13208878e-01 6.95684612e-01 -6.32592738e-01
1.52694464e-01 4.09632564e-01 -1.31071222e+00 1.46862194e-01
1.05484712e+00 7.39134014e-01 -9.76996720e-02 6.35188103e-01
-7.74056539e-02 5.40605783e-01 -1.00880730e+00 -9.27716851e-01
1.77130431e-01 -3.43175888e-01 -9.11862791e-01 -2.96559781e-01
5.45685649e-01 -2.30729699e-01 -5.33122942e-03 1.18650281e+00
4.63895380e-01 1.74547657e-01 1.03365374e+00 -1.06022632e+00
-8.33600998e-01 6.13759220e-01 -4.48496163e-01 3.28673899e-01
7.28460625e-02 5.88063933e-02 6.43747151e-01 -5.52770972e-01
5.79832017e-01 9.17708814e-01 6.14272535e-01 8.27191174e-01
-1.22546196e+00 -7.93309689e-01 3.29875141e-01 5.21384180e-01
-7.91994333e-01 -9.78830680e-02 7.96086788e-01 -5.65396726e-01
4.27678138e-01 -4.52325284e-01 3.32374483e-01 1.44870734e+00
-2.03685865e-01 7.65354872e-01 1.18694353e+00 -6.40710235e-01
5.42819440e-01 7.75029719e-01 1.87042411e-02 1.92459498e-03
3.78379941e-01 4.28654589e-02 -5.33035457e-01 -2.00809613e-01
1.96991637e-01 -2.92338550e-01 -3.86534855e-02 -8.02164495e-01
-6.57768726e-01 1.15595949e+00 1.79097667e-01 3.58651221e-01
-1.33207351e-01 -5.07768810e-01 4.78597283e-01 6.95527971e-01
6.34839654e-01 7.16788173e-01 -9.62234199e-01 -3.03197950e-01
-8.29005599e-01 2.72053450e-01 1.07308626e+00 8.19852769e-01
5.13483167e-01 -1.22486234e-01 3.64749074e-01 8.18982720e-01
7.22470228e-03 4.40497339e-01 8.77045095e-01 -6.82105839e-01
2.88947225e-01 5.87836385e-01 2.46161729e-01 -4.32656705e-01
4.79577892e-02 -5.07185996e-01 -3.72506082e-01 3.81253332e-01
8.41854990e-01 -2.82528251e-01 -8.28638434e-01 1.83441329e+00
7.68064797e-01 4.16462660e-01 3.25186372e-01 4.04505461e-01
5.32931685e-01 2.24708587e-01 1.25881076e-01 -6.95033148e-02
1.04280388e+00 -7.16810822e-01 -2.70826012e-01 -4.19323385e-01
4.61624652e-01 -7.30888009e-01 1.22966218e+00 7.11723447e-01
-8.26676905e-01 -5.04939735e-01 -1.03442657e+00 3.13698292e-01
-7.59187520e-01 -4.31845069e-01 -1.23262860e-01 7.27955997e-01
-6.10204220e-01 4.90910202e-01 -4.57383811e-01 -4.14756805e-01
4.78941351e-01 2.80041873e-01 -3.61049622e-01 -5.27939424e-02
-1.43812799e+00 1.36761653e+00 4.12258059e-01 -8.57540548e-01
-1.02615142e+00 -9.93717492e-01 -4.41904545e-01 1.15192764e-01
3.63655597e-01 -4.20745850e-01 1.82623076e+00 -1.13554299e+00
-1.66699064e+00 8.81660223e-01 3.31400961e-01 -7.39395082e-01
6.36760950e-01 -3.24911892e-01 -4.43396926e-01 6.39435872e-02
2.51195002e-02 9.11351219e-02 1.04878211e+00 -1.14327621e+00
-9.01914299e-01 -3.83604765e-01 -5.61991073e-02 2.00681135e-01
-8.28618288e-01 -1.22052543e-01 2.12616235e-01 -5.27866542e-01
-6.07200086e-01 -8.40957463e-01 -6.43713325e-02 -4.20033962e-01
3.51421505e-01 -5.04676402e-01 1.04035485e+00 -4.87322927e-01
1.13633060e+00 -2.02374554e+00 4.06342894e-02 3.03977907e-01
1.40485570e-01 8.52973282e-01 -1.25091344e-01 3.75161111e-01
-1.37637094e-01 5.55194765e-02 -2.40410671e-01 -4.68357839e-02
-1.01221204e-01 1.75909951e-01 -4.07907277e-01 2.94626117e-01
1.54320940e-01 4.63518411e-01 -1.18893814e+00 -3.08120072e-01
-4.58051637e-02 2.69669861e-01 -4.33982641e-01 5.02595961e-01
-2.26265043e-01 5.22352397e-01 -3.33937973e-01 1.69283777e-01
3.99239779e-01 -1.34565920e-01 1.56861231e-01 3.69036198e-01
4.21363592e-01 5.02028763e-01 -1.20304239e+00 1.49519074e+00
-7.37652540e-01 5.81134915e-01 2.16606990e-01 -1.33378661e+00
1.03289473e+00 3.50052863e-01 2.66304970e-01 -3.13857883e-01
1.42951995e-01 3.86528879e-01 2.17130557e-01 -1.91561222e-01
2.22516164e-01 -4.52963740e-01 -1.16908155e-01 6.96484804e-01
6.56566262e-01 -5.59673429e-01 9.48706716e-02 2.84823656e-01
1.10615015e+00 -9.71196741e-02 7.25866437e-01 -1.36369571e-01
4.99254763e-01 1.75910160e-01 3.49851280e-01 7.42316604e-01
-5.05609691e-01 2.93784767e-01 4.46050763e-01 -3.18664372e-01
-9.73905265e-01 -1.22630155e+00 -7.56209493e-02 1.54519069e+00
-1.57945275e-01 -9.30884778e-02 -8.60819399e-01 -1.53589594e+00
1.56704843e-01 9.82123673e-01 -8.28014910e-01 -5.99631667e-01
-3.51783901e-01 -2.48686537e-01 3.93598378e-01 4.12997007e-01
2.72486866e-01 -7.51006126e-01 -5.06048441e-01 3.89166772e-01
1.46930486e-01 -1.01339948e+00 -1.01532482e-01 3.08664143e-01
-7.22024500e-01 -1.16806638e+00 -5.95136821e-01 -6.01741254e-01
5.35373986e-01 3.40938475e-03 1.44846523e+00 -4.02470559e-01
3.61782730e-01 7.40993679e-01 -5.57969570e-01 -9.82739687e-01
-8.77513409e-01 3.52919430e-01 3.64792049e-01 -2.58291036e-01
1.02704179e+00 -8.43411267e-01 -2.55770922e-01 3.88733715e-01
-1.05140722e+00 -6.03285730e-01 4.55148935e-01 1.07672560e+00
1.88884839e-01 2.02466011e-01 9.45762098e-01 -1.33777368e+00
9.55766559e-01 -7.91278601e-01 -6.03735030e-01 4.50134426e-01
-9.00149167e-01 -6.92158490e-02 1.03892994e+00 -1.07300735e+00
-1.17582393e+00 -7.18863904e-02 1.12211734e-01 -3.69731218e-01
-5.62328219e-01 3.65865350e-01 -1.75803732e-02 -1.35608196e-01
1.25471401e+00 6.63773865e-02 -4.57865410e-02 -4.88823384e-01
4.29856628e-01 7.96064317e-01 6.46214724e-01 -9.03094471e-01
1.10286999e+00 1.06722161e-01 -3.24847877e-01 -3.57783735e-01
-1.18330538e+00 -4.63183880e-01 -1.11174107e+00 5.48277572e-02
2.83650070e-01 -9.25353646e-01 -2.90737510e-01 3.83401990e-01
-9.83155429e-01 -5.12416959e-01 -6.11814022e-01 4.51830536e-01
-4.89717960e-01 1.55367926e-01 -7.35799149e-02 -5.29178143e-01
-6.64086118e-02 -8.87122512e-01 4.86781597e-01 5.00708163e-01
-3.44807625e-01 -1.50331879e+00 4.48199540e-01 2.72390366e-01
4.88346726e-01 -3.51376049e-02 8.42810988e-01 -1.82032990e+00
2.91597486e-01 -2.65641898e-01 2.89822102e-01 7.92322516e-01
1.40338883e-01 -1.42688975e-01 -1.06402004e+00 -2.75010377e-01
2.92891771e-01 -5.69801509e-01 5.78062534e-01 -4.44180928e-02
9.18875992e-01 -3.04070592e-01 -1.60357188e-02 1.45152792e-01
1.05357671e+00 1.23039365e-01 2.34555136e-02 7.52664745e-01
3.72338772e-01 8.51481438e-01 6.40046239e-01 2.60364980e-01
3.26077104e-01 4.29170340e-01 2.38969281e-01 2.54457206e-01
3.18268500e-02 -5.39396048e-01 4.35907423e-01 5.59294462e-01
3.90115947e-01 -9.53971222e-03 -1.26495647e+00 1.08815622e+00
-1.71190202e+00 -8.45733106e-01 3.20940107e-01 2.31028867e+00
1.40015471e+00 4.28339332e-01 4.57911998e-01 -7.14036226e-02
6.31057680e-01 -2.91426480e-01 -9.01621640e-01 -6.44027054e-01
1.51973128e-01 6.08166397e-01 4.12506819e-01 4.89915669e-01
-1.00324082e+00 1.00184596e+00 5.80907822e+00 9.43112969e-01
-1.01786590e+00 3.51127654e-01 4.32508409e-01 -6.74454821e-03
-1.34025842e-01 -1.58905491e-01 -9.22559500e-01 6.17964923e-01
1.39481044e+00 -4.18682009e-01 2.56921262e-01 1.29384649e+00
-4.21078086e-01 1.14546297e-02 -1.10841739e+00 5.24722576e-01
1.16139323e-01 -8.83581042e-01 -2.05392450e-01 -1.38140460e-02
1.07568932e+00 3.06979287e-02 4.01255459e-01 5.77654004e-01
1.20885730e+00 -8.47967267e-01 3.40555131e-01 -1.62471488e-01
4.73129213e-01 -8.96607876e-01 7.81730354e-01 7.10120559e-01
-4.46184814e-01 -2.88580775e-01 -2.30153099e-01 9.24564525e-02
-2.54030794e-01 6.23693466e-01 -1.58887112e+00 1.66763425e-01
8.14936101e-01 5.92557490e-01 -7.26599693e-01 6.77894831e-01
-6.92059159e-01 1.23897111e+00 -4.05065477e-01 2.45595518e-02
1.02890596e-01 1.11527234e-01 4.70006436e-01 1.09819853e+00
1.93493500e-01 -2.32791677e-02 -7.84091130e-02 5.24442494e-01
-2.53807873e-01 1.49571279e-03 -8.90901506e-01 7.50829503e-02
8.05268645e-01 1.05069101e+00 -2.06941769e-01 -4.69725758e-01
-4.28425103e-01 6.60075963e-01 4.71791476e-01 3.34038228e-01
-4.94917244e-01 -5.11020899e-01 7.57730842e-01 1.61399171e-01
4.77865934e-01 -7.89479837e-02 -3.75737280e-01 -1.10608697e+00
-1.63433611e-01 -1.27217245e+00 6.93758786e-01 -3.71307880e-01
-1.85418820e+00 3.85167897e-01 2.63567328e-01 -1.38708568e+00
-6.87708497e-01 -7.23255634e-01 -7.22773671e-01 7.40582228e-01
-1.67881858e+00 -9.92878854e-01 1.80446003e-02 8.94860744e-01
3.88720781e-01 -7.64646471e-01 8.08628380e-01 -9.16342717e-03
-7.97177553e-02 7.95815408e-01 5.21901011e-01 2.75158644e-01
1.38665187e+00 -1.59592366e+00 4.13274854e-01 7.90595293e-01
3.41714978e-01 5.45928717e-01 9.51566041e-01 -5.36712825e-01
-7.89197683e-01 -8.15303922e-01 7.12520421e-01 -1.02186847e+00
1.06030655e+00 -2.60908753e-01 -1.24461436e+00 7.11376011e-01
3.12688917e-01 -5.20909624e-03 1.10576081e+00 3.25627297e-01
-7.77627945e-01 -1.27484262e-01 -1.47216630e+00 3.31933916e-01
4.40963775e-01 -5.43781638e-01 -1.12920868e+00 2.46426553e-01
4.57684100e-01 -4.34650064e-01 -1.06869078e+00 2.41064578e-01
2.21919701e-01 -8.99748147e-01 7.68849790e-01 -1.10674953e+00
4.12872255e-01 -2.91699823e-03 1.19054876e-01 -2.02286720e+00
-8.88875052e-02 -3.89120489e-01 -3.67142200e-01 1.42933929e+00
3.61339450e-01 -6.25154316e-01 7.81423092e-01 5.74444056e-01
3.03567171e-01 -4.66028810e-01 -9.15720403e-01 -8.85313988e-01
8.39293718e-01 -1.33227527e-01 6.59971297e-01 1.33968425e+00
-7.70707428e-03 5.35007179e-01 1.04766347e-01 1.81293726e-01
5.69635451e-01 -3.83682922e-02 1.11121023e+00 -1.66487026e+00
-2.77194917e-01 -4.29944664e-01 -2.11397856e-01 -8.72163832e-01
5.47658861e-01 -5.50381541e-01 -1.74257100e-01 -1.06816387e+00
-1.34216145e-01 -3.23759347e-01 -5.47665536e-01 4.53838557e-01
-3.28368008e-01 -2.08960637e-01 7.00757280e-02 -1.41001746e-01
-5.91319382e-01 4.02107924e-01 9.58346188e-01 -1.07051954e-01
-2.56319165e-01 3.39885145e-01 -1.11431754e+00 8.68365109e-01
9.18709278e-01 -7.26050138e-01 -5.93636394e-01 -2.32553944e-01
4.40477692e-02 -5.45544982e-01 1.10582307e-01 -8.95329475e-01
4.05941784e-01 -4.78630632e-01 4.11518246e-01 4.81012426e-02
1.14937671e-01 -1.10348153e+00 -5.71751177e-01 2.85501152e-01
-4.07735705e-01 -1.77114815e-01 3.16957802e-01 6.50513649e-01
-4.07358795e-01 -5.70377529e-01 1.12504125e+00 -1.66915596e-01
-7.41105139e-01 -1.71212703e-02 -5.94781749e-02 6.20235384e-01
1.15799356e+00 -9.21537578e-02 -5.39443195e-01 -5.85439026e-01
-4.99270946e-01 2.00938344e-01 5.91640651e-01 4.71808195e-01
3.31553251e-01 -1.14962077e+00 -8.59979033e-01 1.95271209e-01
2.10108057e-01 2.87304044e-01 -6.89174384e-02 4.89769459e-01
1.10683213e-04 -2.24051345e-02 -3.48899215e-01 -5.43416560e-01
-1.16437340e+00 5.75031400e-01 2.58120537e-01 -5.67830384e-01
4.57956269e-02 1.03337359e+00 1.45729199e-01 -8.62840533e-01
1.51957870e-01 2.54919846e-02 -2.45863989e-01 2.77792752e-01
6.19611382e-01 1.81967065e-01 2.29405120e-01 -3.43201965e-01
-8.75924751e-02 3.86157960e-01 -4.65182543e-01 -1.72073200e-01
1.36658049e+00 1.42933235e-01 4.65801686e-01 3.93716574e-01
8.21904898e-01 2.61092722e-01 -1.36444533e+00 -6.53824508e-01
1.32182702e-01 -4.16321367e-01 -2.60683835e-01 -1.36156189e+00
-6.46993697e-01 1.18631637e+00 5.14232337e-01 2.92306781e-01
1.14949894e+00 -1.52024422e-02 5.56515694e-01 4.09562171e-01
1.97714001e-01 -1.41151071e+00 3.01678777e-01 7.56709874e-01
4.10679966e-01 -1.54018819e+00 9.85735133e-02 1.76254332e-01
-9.75296617e-01 1.13756549e+00 9.72878158e-01 4.15307879e-02
8.61648023e-01 -2.93938555e-02 2.36369282e-01 2.59937376e-01
-8.93855989e-01 -9.36135501e-02 4.40248340e-01 1.11940646e+00
2.08190739e-01 9.56542641e-02 6.02200860e-03 8.21083844e-01
-3.06969881e-01 -2.78201133e-01 6.32353663e-01 8.93953919e-01
-6.59057617e-01 -1.43390572e+00 -4.37758863e-01 3.01651120e-01
-6.11190975e-01 1.09251671e-01 -9.57649112e-01 8.16194534e-01
1.01661652e-01 9.32997525e-01 -2.99178094e-01 -2.69804001e-01
5.37799001e-01 4.35462356e-01 3.42411935e-01 -1.04800212e+00
-9.64534104e-01 -3.69857162e-01 -1.88665882e-01 3.40298265e-02
-3.53758544e-01 -5.49573123e-01 -7.67545283e-01 -1.51928902e-01
-1.57308757e-01 6.05917811e-01 6.97338700e-01 9.69974339e-01
3.16584975e-01 6.40840977e-02 5.74202001e-01 -2.89908737e-01
-1.23170853e+00 -1.11831546e+00 -5.08607030e-01 7.20041692e-01
4.28128928e-01 -8.19202840e-01 -5.04282475e-01 2.12488636e-01] | [10.80500602722168, 7.6890435218811035] |
1885f1c7-9b41-48ba-bde5-969acbd6ded5 | phone-based-keyword-spotting-for-transcribing | null | null | https://aclanthology.org/2021.alta-1.8 | https://aclanthology.org/2021.alta-1.8.pdf | Phone Based Keyword Spotting for Transcribing Very Low Resource Languages | We investigate the efficiency of two very different spoken term detection approaches for transcription when the available data is insufficient to train a robust speech recognition system. This work is grounded in a very low-resource language documentation scenario where only a few minutes of recording have been transcribed for a given language so far. Experiments on two oral languages show that a pretrained universal phone recognizer, fine-tuned with only a few minutes of target language speech, can be used for spoken term detection through searches in phone confusion networks with a lexicon expressed as a finite state automaton. Experimental results show that a phone recognition based approach provides better overall performances than Dynamic Time Warping when working with clean data, and highlight the benefits of each methods for two types of speech corpus. | ['Laurent Besacier', 'Steven Bird', 'Eric Le Ferrand'] | null | null | null | null | alta-2021-12 | ['robust-speech-recognition', 'keyword-spotting'] | ['speech', 'speech'] | [ 3.29545796e-01 8.70585740e-02 -1.00575007e-01 -4.56606567e-01
-1.33699322e+00 -8.56218934e-01 9.38877404e-01 -3.89206856e-02
-5.98382831e-01 4.94532645e-01 3.19597691e-01 -5.43042600e-01
1.29814282e-01 -1.91893399e-01 -5.65968081e-02 -6.14409149e-01
-5.25691845e-02 7.75862575e-01 2.89884984e-01 -5.31955719e-01
8.04693848e-02 5.63176811e-01 -1.33436656e+00 2.16100961e-01
4.31370258e-01 6.75552487e-01 3.58348668e-01 1.13941860e+00
-4.35949951e-01 5.25735974e-01 -1.06112421e+00 3.55480127e-02
1.58396140e-02 -3.08866203e-01 -9.63779509e-01 3.70468497e-01
1.06354542e-01 4.72516641e-02 -3.67516756e-01 7.81088531e-01
7.09140360e-01 4.35088098e-01 5.05361736e-01 -5.54530740e-01
-9.84629020e-02 6.24143898e-01 5.53514004e-01 7.13237464e-01
7.88828909e-01 1.00452356e-01 5.08138657e-01 -9.95281756e-01
4.41996425e-01 1.29786432e+00 4.36474025e-01 5.25774240e-01
-1.01076019e+00 -1.45230964e-01 8.18536952e-02 -2.04937518e-01
-1.68947947e+00 -1.14172208e+00 3.01210046e-01 -1.81417018e-01
1.83346665e+00 5.74248731e-01 2.25081682e-01 1.31263483e+00
-3.58784497e-02 4.27393794e-01 8.77658129e-01 -9.56318498e-01
3.25693190e-01 5.71097612e-01 4.09568995e-01 6.82844698e-01
-5.22261143e-01 -4.46929829e-03 -5.69269478e-01 -2.17709377e-01
3.81440908e-01 -5.67203581e-01 -3.39255661e-01 2.08116233e-01
-1.29902375e+00 8.09554279e-01 -5.88086188e-01 9.87083793e-01
-2.10529014e-01 -4.58419591e-01 7.00381935e-01 6.99516118e-01
6.86687589e-01 2.71517903e-01 -3.25072259e-01 -6.03513896e-01
-1.16069305e+00 -4.82654035e-01 1.32217717e+00 7.90284574e-01
3.73785645e-01 5.36373794e-01 -2.31069047e-03 1.18266428e+00
-1.97841879e-02 6.65546417e-01 1.14748693e+00 -4.50329989e-01
4.56072748e-01 7.17097744e-02 -6.64836168e-02 -5.61747015e-01
-3.52206588e-01 -1.22573331e-01 -5.64954698e-01 -1.84102923e-01
2.77884930e-01 -2.23119691e-01 -1.16226625e+00 1.36857188e+00
-6.37836605e-02 4.45569009e-02 6.43905103e-01 3.96482587e-01
5.04221499e-01 1.21635616e+00 -4.57595259e-01 -6.70993090e-01
1.32604706e+00 -1.20186043e+00 -1.00484753e+00 -6.71328843e-01
7.88016379e-01 -7.97104180e-01 1.04713035e+00 6.52817667e-01
-1.03939569e+00 -3.59290153e-01 -1.15848398e+00 2.35905766e-01
-7.68738210e-01 1.31360844e-01 3.00191790e-01 1.10382140e+00
-1.28819704e+00 4.82228696e-02 -8.95037174e-01 -9.65141356e-01
-6.08536899e-01 3.99746895e-01 -3.50914955e-01 9.80933756e-02
-1.30769873e+00 1.10652804e+00 5.23836255e-01 1.51613757e-01
-9.97879744e-01 4.44213226e-02 -7.86584556e-01 8.46382380e-02
3.79067361e-01 -8.16740617e-02 1.59103513e+00 -6.91030383e-01
-1.99313128e+00 7.64673948e-01 -3.15202713e-01 -5.80164671e-01
1.56919122e-01 5.67757189e-02 -1.03713310e+00 2.79696584e-01
-2.32872516e-01 2.10595839e-02 9.22143281e-01 -8.45749438e-01
-5.06052732e-01 -2.49357391e-02 -4.22501385e-01 2.27197126e-01
-6.02643549e-01 7.68911481e-01 -6.74179554e-01 -5.22825718e-01
-8.93347412e-02 -8.32827628e-01 -6.59910664e-02 -7.91073918e-01
-1.67426243e-01 -3.48880529e-01 6.90602362e-01 -8.98643136e-01
1.66908062e+00 -1.89074743e+00 1.57365066e-04 4.14064258e-01
-6.94888294e-01 6.01811826e-01 -2.02976480e-01 8.76026034e-01
-4.61109355e-02 2.08983451e-01 -3.24744523e-01 -6.98945343e-01
-7.14889094e-02 6.06992185e-01 -5.56426764e-01 2.99649835e-01
-2.25267962e-01 5.67633390e-01 -6.88295066e-01 -3.39538485e-01
3.18925858e-01 5.76505423e-01 3.64858806e-01 1.98618844e-01
1.49812043e-01 -5.09096459e-02 -1.50089692e-02 5.50993025e-01
-7.41700977e-02 2.76949674e-01 3.21737468e-01 4.24210012e-01
-2.75979877e-01 8.85175228e-01 -1.13671565e+00 1.73527825e+00
-7.64407992e-01 7.72166967e-01 1.83663189e-01 -1.02200997e+00
1.03506696e+00 9.75022376e-01 -1.92742288e-01 -6.12334669e-01
1.39566362e-01 5.10830164e-01 -2.15407923e-01 -5.32755554e-01
3.09033960e-01 -8.34512934e-02 -1.08464621e-01 6.84704483e-01
4.61784214e-01 -3.32194328e-01 1.81396946e-01 1.49411727e-02
1.15754735e+00 -6.27153218e-01 2.15449780e-01 -1.03159077e-01
6.46742880e-01 -4.44215834e-02 -4.16858075e-03 9.68234360e-01
-1.20623060e-01 6.52725160e-01 -2.15916634e-01 -1.21659070e-01
-8.30154002e-01 -6.81150377e-01 1.51076376e-01 1.39061737e+00
-4.97994542e-01 -3.17734212e-01 -8.75444651e-01 -4.52218324e-01
-8.53076279e-01 8.18973005e-01 -2.07757894e-02 -1.64061546e-01
-7.97663450e-01 -4.52067286e-01 1.23868859e+00 2.56701738e-01
1.05567664e-01 -1.18065059e+00 -2.22962856e-01 4.68397558e-01
-3.60599548e-01 -1.28160000e+00 -8.14871788e-01 5.46165645e-01
-6.97316408e-01 -2.80579418e-01 -8.54220867e-01 -1.29727578e+00
2.05671072e-01 2.34232754e-01 9.54280972e-01 -1.25015035e-01
-1.50372274e-02 6.05328381e-01 -3.25537264e-01 -1.61731675e-01
-1.08655179e+00 4.08800662e-01 6.71506345e-01 4.86437604e-02
3.15156311e-01 -4.32350308e-01 8.22000578e-02 3.66320133e-01
-8.90163064e-01 -6.08992457e-01 2.93225020e-01 8.33894253e-01
-6.82246611e-02 -5.27252769e-03 5.03425121e-01 -2.24961102e-01
1.09227037e+00 -1.67083338e-01 -4.32877511e-01 6.04479671e-01
-6.26625419e-01 7.70324916e-02 2.22780392e-01 -7.86407471e-01
-1.17091203e+00 1.55517608e-01 -3.27170521e-01 -2.42507443e-01
-3.79644811e-01 8.29002857e-01 8.79578665e-03 -8.45039356e-03
7.78438210e-01 8.40645611e-01 1.23782545e-01 -6.50232017e-01
1.20546773e-01 1.22362792e+00 8.80542934e-01 -1.26986861e-01
5.05822539e-01 1.39946699e-01 -8.55459630e-01 -1.30778003e+00
-3.52116168e-01 -9.96037602e-01 -6.14328563e-01 -4.70125191e-02
5.94333112e-01 -8.83671880e-01 -2.86658138e-01 5.32304287e-01
-1.22401381e+00 -3.99055749e-01 -1.19967811e-01 7.09128499e-01
-5.20394444e-01 5.94129741e-01 -6.52229726e-01 -1.36007261e+00
-4.31175143e-01 -7.77787447e-01 1.10574412e+00 -2.05109090e-01
-3.97789270e-01 -1.27051163e+00 5.47340214e-01 1.46546721e-01
4.16430116e-01 -5.97722113e-01 6.24776959e-01 -1.26081741e+00
4.12275642e-02 -4.81530517e-01 5.37670553e-01 5.83173573e-01
3.44836950e-01 -1.64514501e-03 -1.32150066e+00 -3.83799613e-01
2.65946776e-01 -2.64826357e-01 5.92708588e-01 2.88634356e-02
2.23463744e-01 -6.04217708e-01 -3.09124798e-01 1.24943200e-02
9.26567972e-01 6.79991424e-01 4.76289749e-01 1.72428846e-01
3.54788095e-01 6.66571319e-01 4.19136494e-01 5.60809150e-02
6.17294386e-02 8.33135605e-01 -3.17145646e-01 6.72233105e-02
6.31156489e-02 2.49542948e-02 9.52703059e-01 1.52721930e+00
2.74263382e-01 -8.68560791e-01 -1.41587496e+00 9.85969484e-01
-1.66918373e+00 -8.68854105e-01 4.84421581e-01 2.22366595e+00
9.56401229e-01 2.38725647e-01 1.28467694e-01 3.33707631e-01
6.74904585e-01 1.93037942e-01 5.25421239e-02 -8.81844461e-01
-3.64110589e-01 3.41325879e-01 3.64177078e-01 9.69868064e-01
-8.32409620e-01 1.08106089e+00 7.62953568e+00 1.02506196e+00
-1.37809205e+00 2.67289072e-01 2.03037262e-01 1.38461273e-02
1.39909223e-01 -1.96655348e-01 -7.29257405e-01 2.65338719e-01
2.04919887e+00 -1.44337118e-01 4.45520222e-01 6.20893538e-01
3.91890705e-01 -6.51130676e-02 -1.05996084e+00 1.26961493e+00
3.99536401e-01 -1.13461447e+00 -1.17255382e-01 -7.08011910e-02
3.37980747e-01 4.05497819e-01 -1.99543610e-01 4.33136553e-01
1.96659982e-01 -1.03283286e+00 6.04685307e-01 4.51750576e-01
6.27501369e-01 -7.71728158e-01 6.35090172e-01 7.79276669e-01
-1.15879035e+00 1.97271898e-01 -2.00186357e-01 8.19593146e-02
4.80981559e-01 2.12217271e-01 -1.27953708e+00 3.32818329e-01
3.98117244e-01 7.52839819e-02 -2.71458566e-01 7.35950947e-01
1.62637159e-01 1.06184828e+00 -9.16508317e-01 -3.93749215e-02
4.77021635e-01 -6.34441059e-03 7.40408421e-01 1.75574780e+00
5.59933841e-01 4.11443338e-02 2.59369880e-01 -5.62755857e-03
3.83551359e-01 2.36026376e-01 -7.89971530e-01 -3.98406774e-01
4.65572029e-01 1.03690195e+00 -8.77696395e-01 -5.54429054e-01
-2.29461670e-01 1.35266232e+00 1.89831201e-02 3.61950040e-01
-2.74248928e-01 -5.19264817e-01 3.29585344e-01 -3.63608301e-01
3.46560895e-01 -6.28158927e-01 1.97925717e-01 -1.11934507e+00
1.25002628e-02 -1.19232941e+00 2.31444806e-01 -8.48246753e-01
-8.10446441e-01 1.13878226e+00 -7.86878616e-02 -9.17621672e-01
-1.07041240e+00 -5.15512764e-01 -6.24446034e-01 1.02254307e+00
-9.50275183e-01 -8.21990013e-01 4.05483991e-01 5.57496369e-01
1.09064901e+00 -5.03082275e-01 1.39112902e+00 1.39841706e-01
-6.28559530e-01 6.25938296e-01 2.36688554e-01 3.32438946e-02
5.30086458e-01 -1.04027343e+00 6.25793576e-01 9.79614854e-01
7.27383018e-01 7.20278323e-01 7.57876456e-01 -3.47130448e-01
-1.26134598e+00 -6.27551436e-01 1.55253279e+00 -5.42070031e-01
7.38610923e-01 -7.88407981e-01 -1.28847945e+00 5.17906547e-01
6.05146825e-01 -2.86672592e-01 5.78609347e-01 -3.45290825e-02
-1.24323502e-01 1.40895918e-01 -7.87854850e-01 3.35079104e-01
7.56410480e-01 -1.30771399e+00 -8.66914153e-01 6.77513182e-01
7.21342564e-01 -1.87460810e-01 -7.85200179e-01 -9.77988392e-02
3.30410898e-01 -3.47105294e-01 7.81467974e-01 -4.16896582e-01
-6.89427018e-01 -7.51650601e-04 -1.68286979e-01 -1.33157754e+00
3.14423949e-01 -1.22402000e+00 1.05617205e-02 1.39292002e+00
6.95518255e-01 -7.56493151e-01 2.71757454e-01 2.86256999e-01
-2.32142180e-01 -1.73181012e-01 -1.26829994e+00 -1.17992902e+00
-3.22998911e-01 -6.54839635e-01 2.87167430e-01 8.44275057e-01
3.33955646e-01 7.72994280e-01 -4.20246601e-01 2.89518923e-01
-2.45832667e-01 -3.62311661e-01 3.99290204e-01 -9.84935403e-01
-1.70307934e-01 -1.91559702e-01 -2.68730581e-01 -1.07832587e+00
3.27418089e-01 -5.67729890e-01 6.09325409e-01 -1.30648482e+00
-3.60249162e-01 -7.38700926e-02 -2.41166830e-01 5.81951678e-01
4.15437013e-01 1.96097400e-02 -1.15175143e-01 1.41592711e-01
-2.68908709e-01 3.05250525e-01 1.85493782e-01 -3.10418814e-01
-3.72296929e-01 1.80842042e-01 -1.22053646e-01 5.71097851e-01
6.73316240e-01 -4.38211232e-01 -5.33443570e-01 -3.27683002e-01
-1.15280315e-01 3.43796074e-01 -9.10036415e-02 -1.10533798e+00
4.96675849e-01 1.94456100e-01 -5.61800778e-01 -5.17811775e-01
6.17302835e-01 -6.64228439e-01 1.48859620e-02 2.45478511e-01
-3.70851308e-01 1.39946893e-01 3.78899127e-01 4.59844589e-01
-5.80379307e-01 -4.51426208e-01 4.11750555e-01 -9.61633027e-02
-7.36731529e-01 -2.04592049e-01 -1.31533301e+00 -2.10083544e-01
4.28626060e-01 -4.21873599e-01 5.94457202e-02 -8.07304561e-01
-1.02763641e+00 -1.19713433e-01 -5.36191016e-02 7.58776009e-01
4.23467934e-01 -9.45216656e-01 -4.17122185e-01 2.07344159e-01
1.48613021e-01 -6.78417861e-01 -1.85520738e-01 5.40204763e-01
-8.31978023e-02 1.06499422e+00 3.23563874e-01 -6.67436123e-01
-1.63612187e+00 5.94831824e-01 5.57253063e-01 -2.95705646e-01
-3.74434650e-01 7.71780670e-01 -2.69977480e-01 -5.89544594e-01
6.51317596e-01 -6.78561389e-01 -2.38718331e-01 2.97727585e-01
7.18427658e-01 2.05171749e-01 6.66927993e-01 -9.03347015e-01
-5.19525945e-01 2.45772496e-01 -1.02459386e-01 -9.32028949e-01
9.10508752e-01 -5.91151893e-01 2.09170699e-01 1.09402668e+00
1.19376433e+00 4.61080745e-02 -5.08198977e-01 -4.13291663e-01
3.26509714e-01 2.05970690e-01 2.45882198e-01 -7.60546863e-01
-4.31767821e-01 8.00349116e-01 8.80150497e-01 5.68049550e-01
9.79822874e-01 6.56328723e-03 5.95111012e-01 1.05247724e+00
3.91883612e-01 -1.26186085e+00 -2.15973333e-01 9.55953479e-01
9.39372003e-01 -1.12608325e+00 -6.27814353e-01 -3.08506727e-01
-5.80575347e-01 1.22855341e+00 -6.64310753e-02 3.11890155e-01
5.77766240e-01 3.28985929e-01 5.67765534e-01 -1.69938102e-01
-8.61135602e-01 -3.75289381e-01 1.96624860e-01 5.61062694e-01
9.25045367e-03 -2.05690771e-01 2.61056125e-02 3.20517123e-01
-2.84618139e-01 -3.82514149e-01 5.32471418e-01 1.19815123e+00
-6.36283398e-01 -1.12635052e+00 -4.55467284e-01 -8.33971947e-02
-4.48293090e-01 -3.69393438e-01 -5.71599841e-01 5.66609859e-01
-4.99291658e-01 1.34584522e+00 1.28918231e-01 -1.74890473e-01
3.06582719e-01 8.30279589e-01 1.02877624e-01 -1.03739595e+00
-9.15402830e-01 4.14583355e-01 5.57292759e-01 -4.73389685e-01
-4.80287343e-01 -6.50373757e-01 -1.06430197e+00 2.08773211e-01
-4.96906817e-01 5.25003552e-01 8.82599056e-01 1.18561041e+00
9.62922126e-02 2.37657562e-01 6.79812908e-01 -7.12347329e-01
-5.81510842e-01 -1.24955559e+00 -4.78333026e-01 -4.36401099e-01
6.18836701e-01 -1.66402221e-01 -5.61583936e-01 2.53158987e-01] | [14.370945930480957, 6.711470603942871] |
51fa4293-c254-497b-8888-c366181f8242 | visual-affordance-and-function-understanding | 1807.06775 | null | http://arxiv.org/abs/1807.06775v1 | http://arxiv.org/pdf/1807.06775v1.pdf | Visual Affordance and Function Understanding: A Survey | Nowadays, robots are dominating the manufacturing, entertainment and
healthcare industries. Robot vision aims to equip robots with the ability to
discover information, understand it and interact with the environment. These
capabilities require an agent to effectively understand object affordances and
functionalities in complex visual domains. In this literature survey, we first
focus on Visual affordances and summarize the state of the art as well as open
problems and research gaps. Specifically, we discuss sub-problems such as
affordance detection, categorization, segmentation and high-level reasoning.
Furthermore, we cover functional scene understanding and the prevalent
functional descriptors used in the literature. The survey also provides
necessary background to the problem, sheds light on its significance and
highlights the existing challenges for affordance and functionality learning. | ['Mohammed Hassanin', 'Salman Khan', 'Murat Tahtali'] | 2018-07-18 | null | null | null | null | ['affordance-detection'] | ['computer-vision'] | [ 2.68777370e-01 2.10067660e-01 -5.65816045e-01 -2.90839970e-01
3.36960196e-01 -7.58719921e-01 3.84826213e-01 4.16914344e-01
-1.83234252e-02 4.26392913e-01 4.88155931e-02 8.83637145e-02
-5.16526699e-01 -3.37775677e-01 -6.42935395e-01 -5.27209342e-01
-3.91075671e-01 3.92034531e-01 7.31914714e-02 -3.58813167e-01
6.22611940e-01 8.50747526e-01 -2.15551710e+00 6.10227846e-02
9.10316408e-01 1.16335738e+00 9.12119806e-01 5.17113090e-01
1.01494677e-01 7.11493671e-01 -4.05027360e-01 2.30712891e-01
1.41346663e-01 1.56812087e-01 -8.83891165e-01 6.90305352e-01
2.92461723e-01 -4.38294470e-01 -3.67688417e-01 1.05068719e+00
-2.54160523e-01 3.16883564e-01 9.99617219e-01 -1.91691256e+00
-9.65327501e-01 2.01647118e-01 -7.65324757e-02 -1.48210153e-02
7.33958721e-01 3.11801970e-01 1.26264441e+00 -7.65076518e-01
8.59669924e-01 1.42771626e+00 1.04313225e-01 5.86798370e-01
-9.13815260e-01 1.42373487e-01 5.15382469e-01 5.32232940e-01
-8.30842316e-01 -1.02135487e-01 6.09096885e-01 -6.44719243e-01
1.32493293e+00 2.86494464e-01 1.11990845e+00 6.90013170e-01
1.31022647e-01 1.31690073e+00 7.41103053e-01 -4.42992955e-01
1.47056669e-01 -3.93672615e-01 4.18248355e-01 9.09357905e-01
4.80501115e-01 2.23990321e-01 -3.31148326e-01 2.87302405e-01
1.24874282e+00 3.36300790e-01 -2.90732354e-01 -1.23691356e+00
-1.78229678e+00 7.05268621e-01 9.60454524e-01 4.23799902e-01
-6.15458965e-01 4.18051094e-01 2.35267788e-01 1.96685597e-01
-4.05779243e-01 1.08135355e+00 -6.66063607e-01 2.10997567e-01
1.52766317e-01 3.04019302e-01 7.76953757e-01 1.50620925e+00
7.92223454e-01 1.87248997e-02 1.55567005e-01 6.35662794e-01
4.23295081e-01 3.18087876e-01 1.16937071e-01 -1.38820148e+00
1.08835518e-01 7.58781195e-01 1.94900170e-01 -8.41272473e-01
-8.79456401e-01 4.41093534e-01 -1.84310541e-01 4.27855879e-01
2.93939024e-01 4.10172582e-01 -1.05734205e+00 1.24485016e+00
3.48526001e-01 -4.79846418e-01 5.06355762e-02 1.25746655e+00
1.05657160e+00 2.75591850e-01 3.22501391e-01 2.00755540e-02
1.55647266e+00 -1.40059662e+00 -8.23328912e-01 -2.47446299e-01
4.33850169e-01 -4.29503441e-01 1.25695372e+00 2.38620624e-01
-5.87846637e-01 -7.10281193e-01 -1.33957613e+00 -7.56896973e-01
-6.62573278e-01 1.88313499e-01 1.54958522e+00 8.79331231e-02
-6.83652520e-01 5.39633155e-01 -8.37326467e-01 -7.82183886e-01
5.78909218e-01 5.55431783e-01 -6.74336672e-01 -2.73524195e-01
-5.97985566e-01 1.24533796e+00 7.54300892e-01 -1.89136446e-01
-1.18299735e+00 -4.35487986e-01 -1.47230124e+00 -2.64823049e-01
8.20947766e-01 -9.07397509e-01 1.47719657e+00 -5.90626359e-01
-1.19092631e+00 8.42006326e-01 8.07432681e-02 -1.77784562e-01
1.67584166e-01 -3.94907355e-01 -2.69737747e-02 3.75938058e-01
3.31722796e-01 1.13405037e+00 5.60611308e-01 -1.53519499e+00
-1.02606320e+00 -4.52111870e-01 7.87576318e-01 5.70596278e-01
2.61171162e-01 -3.62866879e-01 -1.79541230e-01 -3.00855309e-01
6.27808213e-01 -8.14421833e-01 -2.64601588e-01 4.15471941e-01
-5.21629930e-01 -6.57514095e-01 8.29128087e-01 -2.49983147e-01
3.82294625e-01 -2.04016900e+00 5.70406914e-01 -2.46997431e-01
5.41153848e-01 -4.81446654e-01 3.31928283e-02 5.86850286e-01
5.30287996e-02 -2.17738062e-01 -4.84978259e-02 1.29376035e-02
3.05475980e-01 7.39532769e-01 -1.21178366e-02 5.94513118e-01
1.37769818e-01 9.99172151e-01 -1.20792902e+00 -3.03361714e-01
1.00511503e+00 8.50263983e-02 -2.94041872e-01 1.91961169e-01
-5.19598663e-01 7.33074188e-01 -6.86930418e-01 1.16576004e+00
2.08813712e-01 -2.89831161e-02 1.68421566e-01 -2.22080231e-01
-2.93562174e-01 2.32541695e-01 -9.92545903e-01 1.97387922e+00
-5.21819174e-01 7.10666001e-01 8.62390772e-02 -1.04937911e+00
5.57353914e-01 1.35403961e-01 4.21670198e-01 -4.49498624e-01
2.36383945e-01 3.83952469e-01 1.61103293e-01 -1.00483656e+00
4.97858703e-01 8.99410620e-02 -3.39813679e-01 -7.04338327e-02
1.63377807e-01 -7.74728954e-01 4.05265182e-01 -2.88942873e-01
6.91923738e-01 5.22891998e-01 1.01036859e+00 -2.32241377e-01
3.93790603e-01 5.95600605e-01 -9.63632688e-02 4.72685546e-01
-3.46485227e-01 1.33680567e-01 2.70427734e-01 -6.44845963e-01
-9.11522448e-01 -1.23532081e+00 -2.59661134e-02 1.37019539e+00
9.79727030e-01 4.76428941e-02 -5.65137029e-01 -5.88331163e-01
3.33306134e-01 6.51411712e-01 -7.94730365e-01 7.63219967e-03
-4.89192992e-01 -1.25477090e-02 -2.38962725e-01 8.37532580e-01
2.93253422e-01 -1.54753125e+00 -1.41581833e+00 -2.31764898e-01
-1.04898132e-01 -1.26758575e+00 -6.78365082e-02 2.55755156e-01
-9.50493932e-01 -1.47305000e+00 -2.35156685e-01 -1.19329512e+00
8.12903404e-01 7.93613732e-01 1.02308297e+00 5.15306480e-02
-6.38286293e-01 7.59505570e-01 -4.70091909e-01 -4.51563686e-01
-1.95159316e-01 -1.14254281e-01 1.47743046e-01 -5.85851073e-01
2.50510156e-01 -3.51392567e-01 -3.88053447e-01 1.89186603e-01
-6.64248765e-01 1.30678996e-01 7.49148190e-01 6.51413560e-01
7.65099525e-01 -1.01802848e-01 2.68833071e-01 -3.02631080e-01
4.50788110e-01 -3.21151793e-01 -3.84695500e-01 2.10311010e-01
-2.75830716e-01 1.39906546e-02 1.19609050e-01 -3.87563199e-01
-7.82992899e-01 2.72107542e-01 3.48208904e-01 -2.08450466e-01
-7.53477573e-01 1.80480435e-01 -6.52722478e-01 -1.33955702e-01
4.84834045e-01 -1.10679358e-01 -2.26826251e-01 -5.25462031e-01
9.07282352e-01 5.91903508e-01 4.54320252e-01 -5.07112145e-01
1.32578641e-01 7.87774980e-01 1.57004476e-01 -1.11066496e+00
-7.93254375e-01 -8.56670380e-01 -1.28688788e+00 -2.49255732e-01
1.06168878e+00 -4.76677209e-01 -1.06757331e+00 1.10550299e-01
-1.32430851e+00 -2.23960921e-01 -7.02483833e-01 4.81797844e-01
-1.37819552e+00 1.17357917e-01 -4.49395031e-01 -6.79197907e-01
1.44996732e-01 -1.36246800e+00 1.27218902e+00 2.06442729e-01
-4.63629097e-01 -7.82191753e-01 -6.12446904e-01 2.91822612e-01
-2.50825614e-01 6.09917045e-01 1.12706292e+00 -4.22177672e-01
-8.79433453e-01 5.02655953e-02 -4.41637695e-01 -1.25855242e-03
5.07310987e-01 -5.75296104e-01 -5.85433602e-01 1.78539932e-01
4.62439209e-02 -4.34699148e-01 5.92418492e-01 5.11209249e-01
9.21323657e-01 4.49585437e-04 -6.98397517e-01 2.87471324e-01
1.28567183e+00 4.35429871e-01 2.55048394e-01 3.79317671e-01
1.05270123e+00 1.34319866e+00 1.18399644e+00 1.05552107e-01
5.72441876e-01 4.24957126e-01 1.06098151e+00 2.63990492e-01
-1.45288244e-01 -1.73124358e-01 -3.51321369e-01 2.53908038e-01
-3.48375767e-01 2.54555643e-02 -7.03523874e-01 4.16713625e-01
-2.11736393e+00 -4.46130902e-01 -2.89743692e-01 1.55781245e+00
2.27240190e-01 -1.72849789e-01 4.91221584e-02 2.99654007e-01
5.55947602e-01 -1.67432100e-01 -9.97549772e-01 -2.71638870e-01
7.71585777e-02 -3.92740637e-01 4.21035767e-01 2.37426430e-01
-1.49447906e+00 1.21600568e+00 6.86607122e+00 1.92331254e-01
-6.60159051e-01 -1.74890440e-02 -8.86152238e-02 4.77431595e-01
1.88229829e-01 -1.18714765e-01 -2.06887439e-01 -3.59622329e-01
-2.25297771e-02 2.98165232e-01 6.78810537e-01 1.35836279e+00
-7.36433193e-02 -5.24303436e-01 -1.82222342e+00 1.17782116e+00
2.46205166e-01 -9.94488001e-01 1.12081185e-01 1.26884818e-01
5.25224924e-01 -8.38322639e-02 -8.80220383e-02 -3.39076780e-02
1.38361290e-01 -1.18103397e+00 1.20975113e+00 5.14245689e-01
6.04584277e-01 -5.06467462e-01 3.29524636e-01 3.65424365e-01
-1.35839021e+00 -6.48243129e-01 -3.75498503e-01 -5.84750831e-01
2.47556031e-01 -1.39747366e-01 -7.16304600e-01 5.47256291e-01
7.00615466e-01 1.09658325e+00 -9.13563818e-02 1.02359605e+00
-6.48135304e-01 -7.03541279e-01 -5.72421812e-02 -4.74275231e-01
1.57644108e-01 -1.74201891e-01 5.57974994e-01 7.12833583e-01
-1.01712339e-01 -6.53551593e-02 5.77006578e-01 8.73478651e-01
3.04025114e-01 -1.81383103e-01 -7.95044243e-01 -2.01236159e-01
2.95953035e-01 1.04482722e+00 -1.08211708e+00 -3.21587622e-01
-4.96219367e-01 1.13154471e+00 2.81858355e-01 2.61698276e-01
-5.29912055e-01 -4.49592799e-01 1.12831223e+00 -1.37473419e-01
1.00301355e-01 -8.55339825e-01 -5.36144912e-01 -7.00505316e-01
-5.68683483e-02 -6.55537367e-01 1.18439309e-01 -1.00056458e+00
-1.09149051e+00 1.06043205e-01 4.02855068e-01 -1.25509989e+00
-6.95215017e-02 -1.48767412e+00 3.26638073e-01 2.63042986e-01
-1.54590309e+00 -1.37177539e+00 -4.93646652e-01 3.80080909e-01
1.18838131e+00 2.15444222e-01 8.51674676e-01 -3.65522861e-01
2.64305681e-01 -4.74433094e-01 -4.60631907e-01 -1.32049888e-01
-1.02030762e-01 -1.50939965e+00 2.80844986e-01 1.33671835e-01
1.78998575e-01 7.87663341e-01 8.35254669e-01 -4.57608521e-01
-1.80287123e+00 -4.64507282e-01 3.91390741e-01 -7.28177130e-01
6.57871008e-01 -3.55912954e-01 -3.99303645e-01 7.51641154e-01
8.51172209e-02 6.67237043e-02 3.72104406e-01 -1.32372871e-01
-2.27538377e-01 4.40693438e-01 -1.33635509e+00 8.47171783e-01
1.68924010e+00 -3.17396939e-01 -1.13055301e+00 4.13327813e-01
1.01841402e+00 -5.70274413e-01 -6.66115165e-01 3.81379366e-01
9.43812430e-01 -9.05794203e-01 1.08682895e+00 -6.79059148e-01
7.28076100e-02 -4.08934742e-01 -5.28213799e-01 -1.07223606e+00
-3.19724619e-01 -1.57721981e-01 -3.98698688e-01 3.68163049e-01
-1.62700750e-02 -5.19577444e-01 3.49374086e-01 3.97830784e-01
-6.09473348e-01 -5.09912312e-01 -6.35731041e-01 -7.07692444e-01
-5.37051618e-01 -6.60322070e-01 6.19260192e-01 7.29725838e-01
4.55412626e-01 4.94472235e-01 1.07684344e-01 2.06122950e-01
5.15666723e-01 4.71740723e-01 5.70903659e-01 -1.71082318e+00
1.90505937e-01 -5.89472294e-01 -7.44821489e-01 -1.38323009e+00
2.87920386e-01 -6.84134126e-01 4.09107119e-01 -2.31834173e+00
-2.84733139e-02 -2.65179008e-01 6.61118776e-02 5.21477401e-01
3.17490757e-01 8.73927698e-02 4.18892771e-01 2.41924435e-01
-7.34672546e-01 3.46262246e-01 2.03890061e+00 -2.94151455e-01
-2.48515621e-01 -1.51847795e-01 -7.10969090e-01 9.93812621e-01
7.75099337e-01 -4.09340523e-02 -6.09998703e-01 -3.85062099e-01
7.15718716e-02 -2.10916787e-01 4.50383425e-01 -7.73636162e-01
-1.71704769e-01 -4.66632754e-01 3.09367299e-01 -4.91493583e-01
6.18279159e-01 -1.21183372e+00 -5.23061037e-01 7.56508112e-01
-8.61197188e-02 9.99570861e-02 -9.45073590e-02 6.92575514e-01
1.26894685e-02 -3.66897345e-01 4.49433088e-01 -5.65449059e-01
-1.71065462e+00 1.37664124e-01 -5.17604291e-01 -4.71190631e-01
1.42692292e+00 -5.48093617e-01 -1.45154178e-01 -2.72727460e-02
-1.00214672e+00 6.05383217e-01 5.69253922e-01 1.04024649e+00
8.31242621e-01 -1.06199348e+00 5.26663437e-02 1.63077727e-01
6.86984658e-01 8.96714106e-02 2.70148933e-01 5.62796295e-01
-7.98826098e-01 6.36726022e-01 -4.99554724e-01 -7.56048024e-01
-9.29812312e-01 1.16303694e+00 3.46808024e-02 4.99927998e-01
-6.10706687e-01 6.24621809e-01 6.31537795e-01 -6.96781456e-01
4.39873725e-01 -1.07741547e+00 -7.09510863e-01 -3.22220057e-01
1.95305735e-01 6.60201192e-01 -3.71738255e-01 -9.78135467e-01
-3.99206221e-01 7.39423692e-01 6.41608953e-01 1.40614271e-01
1.00399518e+00 -4.77467626e-01 -4.19807166e-01 8.99235487e-01
9.28877234e-01 -8.25534940e-01 -1.42380524e+00 2.98947960e-01
4.77903485e-01 -3.41578901e-01 -3.52391839e-01 -4.98356462e-01
-4.37058926e-01 9.83952880e-01 4.32165772e-01 4.60417360e-01
8.25663030e-01 7.62839437e-01 3.55605870e-01 8.87735486e-01
8.83905292e-01 -1.10378110e+00 6.44366682e-01 6.52447402e-01
1.35885310e+00 -1.31479681e+00 2.60969129e-04 -1.08132756e+00
-7.35800207e-01 1.18105388e+00 6.88761950e-01 -3.11083734e-01
6.39998972e-01 4.50694934e-02 -1.53769210e-01 -4.74624783e-01
-1.62081838e-01 -7.76001930e-01 3.92852515e-01 1.37609839e+00
7.57207647e-02 3.54642630e-01 3.18770632e-02 9.02582631e-02
-3.73381406e-01 -3.12652230e-01 2.75712669e-01 1.22865510e+00
-7.97895312e-01 -6.57853782e-01 -1.79966286e-01 3.64374161e-01
8.88153687e-02 2.49371067e-01 -2.97974408e-01 1.05077612e+00
5.21859169e-01 1.13258922e+00 2.16970429e-01 -1.41203836e-01
6.80002511e-01 -2.87061661e-01 1.06734717e+00 -9.51660037e-01
1.11345328e-01 -1.08774759e-01 1.33222580e-01 -7.89046645e-01
-8.41150880e-01 -7.50512660e-01 -1.84895623e+00 4.66562450e-01
-3.53432566e-01 -3.29639345e-01 9.38739061e-01 1.23243487e+00
-6.43512830e-02 8.67750108e-01 -4.50507887e-02 -1.45416522e+00
-7.04850927e-02 -9.82583284e-01 -6.43674910e-01 1.15055248e-01
6.68372035e-01 -1.36941457e+00 -4.62032743e-02 9.63995606e-02] | [5.154612064361572, -0.09053874015808105] |
e9221826-c4e0-4d9f-a6b4-809aa0147702 | pp-matting-high-accuracy-natural-image | 2204.09433 | null | https://arxiv.org/abs/2204.09433v1 | https://arxiv.org/pdf/2204.09433v1.pdf | PP-Matting: High-Accuracy Natural Image Matting | Natural image matting is a fundamental and challenging computer vision task. It has many applications in image editing and composition. Recently, deep learning-based approaches have achieved great improvements in image matting. However, most of them require a user-supplied trimap as an auxiliary input, which limits the matting applications in the real world. Although some trimap-free approaches have been proposed, the matting quality is still unsatisfactory compared to trimap-based ones. Without the trimap guidance, the matting models suffer from foreground-background ambiguity easily, and also generate blurry details in the transition area. In this work, we propose PP-Matting, a trimap-free architecture that can achieve high-accuracy natural image matting. Our method applies a high-resolution detail branch (HRDB) that extracts fine-grained details of the foreground with keeping feature resolution unchanged. Also, we propose a semantic context branch (SCB) that adopts a semantic segmentation subtask. It prevents the detail prediction from local ambiguity caused by semantic context missing. In addition, we conduct extensive experiments on two well-known benchmarks: Composition-1k and Distinctions-646. The results demonstrate the superiority of PP-Matting over previous methods. Furthermore, we provide a qualitative evaluation of our method on human matting which shows its outstanding performance in the practical application. The code and pre-trained models will be available at PaddleSeg: https://github.com/PaddlePaddle/PaddleSeg. | ['dianhai yu', 'Xiaoguang Hu', 'Qingqing Dang', 'Yuning Du', 'Zhiliang Yu', 'Zeyu Chen', 'Zewu Wu', 'Shiyu Tang', 'Lutao Chu', 'Yuying Hao', 'Juncai Peng', 'Jian Wang', 'Yi Liu', 'Guowei Chen'] | 2022-04-20 | null | null | null | null | ['image-matting'] | ['computer-vision'] | [ 3.61635387e-01 -1.39954343e-01 -2.85371877e-02 -3.28448921e-01
-6.58629358e-01 -1.60400093e-01 5.39215744e-01 -4.01475847e-01
-1.46528482e-01 4.10715163e-01 1.20351296e-02 -3.19124818e-01
3.19138497e-01 -7.96027720e-01 -9.18631256e-01 -8.24469864e-01
5.53459585e-01 2.82411456e-01 5.02048790e-01 -1.99643642e-01
7.48451278e-02 1.09756924e-01 -1.30792642e+00 5.88567078e-01
1.36729205e+00 9.72873867e-01 6.57650530e-01 3.30647260e-01
-5.52330077e-01 4.89549190e-01 -3.95770401e-01 -5.18380582e-01
4.90192115e-01 -4.74425107e-01 -4.84067380e-01 4.39816117e-01
7.11582303e-01 -3.84428650e-01 -4.21634376e-01 1.24497414e+00
1.91551790e-01 -1.60053149e-01 3.66125345e-01 -1.24425483e+00
-6.62195086e-01 4.02082920e-01 -9.12044108e-01 2.22133715e-02
1.54807735e-02 5.52727759e-01 5.82629323e-01 -1.04478633e+00
3.22012961e-01 1.33855689e+00 4.94203240e-01 5.52311063e-01
-1.12663734e+00 -8.42630267e-01 3.63934159e-01 3.07351977e-01
-1.08268607e+00 -3.99071246e-01 8.23217511e-01 -3.55465084e-01
3.35464478e-01 3.64947677e-01 6.71798706e-01 9.11931634e-01
3.33700657e-01 1.04550648e+00 1.34380782e+00 -1.87535971e-01
1.81532744e-02 -1.94933727e-01 -1.36662304e-01 7.68126965e-01
1.65389165e-01 -1.55577287e-01 -4.40415323e-01 2.33881846e-01
1.14329684e+00 2.74082810e-01 -5.58752537e-01 -2.63992965e-01
-1.43908441e+00 4.55024779e-01 6.38245583e-01 9.08347666e-02
-2.94625223e-01 1.47574112e-01 1.99719265e-01 1.94981501e-01
6.12902641e-01 1.44272536e-01 -2.17747763e-01 1.26829579e-01
-1.17133093e+00 2.34029919e-01 2.65965253e-01 1.18896675e+00
8.02535832e-01 1.52965233e-01 -3.99833620e-01 9.83372986e-01
1.16473489e-01 6.34734988e-01 2.97847271e-01 -1.00745916e+00
7.51422763e-01 6.75283492e-01 1.08970888e-01 -1.08285439e+00
-5.44469915e-02 -3.30897957e-01 -1.24634635e+00 2.44016960e-01
4.75135714e-01 1.10086598e-01 -1.28736782e+00 1.36959076e+00
3.97212535e-01 3.71140242e-01 -2.18495443e-01 1.22475994e+00
9.19711411e-01 9.67152894e-01 -1.07622355e-01 -8.87616873e-02
1.38837481e+00 -1.58156216e+00 -8.71279478e-01 -4.47710276e-01
6.32833242e-02 -1.09302950e+00 1.45709896e+00 5.74914873e-01
-1.21194851e+00 -6.59458756e-01 -9.23748910e-01 -2.64797211e-01
-6.36973232e-02 2.66760707e-01 6.62233412e-01 3.61207277e-01
-8.82914007e-01 4.56493586e-01 -9.60688293e-01 -5.75747304e-02
6.10684633e-01 7.07138851e-02 -2.51993239e-01 -4.30424333e-01
-8.77583802e-01 5.23099482e-01 3.58724385e-01 5.11647344e-01
-1.04323220e+00 -5.80911517e-01 -9.36235368e-01 -7.40263611e-02
7.01252639e-01 -7.76166975e-01 1.20149028e+00 -1.11005008e+00
-1.45245636e+00 6.60238445e-01 -3.38509053e-01 -1.91064283e-01
8.04484665e-01 -4.24127430e-01 -2.78175652e-01 2.87140571e-02
1.37947693e-01 6.99320495e-01 1.17860055e+00 -1.38675416e+00
-5.41569054e-01 -1.11312583e-01 8.71224701e-02 2.18469039e-01
-2.16705665e-01 -1.79821685e-01 -9.14315343e-01 -1.22724378e+00
2.51700342e-01 -6.70191765e-01 -2.77945846e-01 1.71870366e-01
-5.18954158e-01 2.27749169e-01 9.62955475e-01 -9.30892766e-01
1.18943000e+00 -2.18134713e+00 2.12805673e-01 -2.07800403e-01
2.71232307e-01 5.13560534e-01 -3.57623309e-01 2.05348328e-01
9.81207639e-02 -2.50337925e-02 -6.05180681e-01 -5.16078651e-01
-2.08729520e-01 2.41301417e-01 -4.29727316e-01 1.21011682e-01
2.96140581e-01 1.03524745e+00 -7.41709232e-01 -5.57278633e-01
4.47956771e-01 3.42974037e-01 -4.53678221e-01 3.33536327e-01
-5.61744690e-01 5.48820794e-01 -2.70763248e-01 7.14624166e-01
1.14814436e+00 -2.41290197e-01 -1.34147620e-02 -4.62248176e-01
-5.95873445e-02 1.09298453e-01 -1.12635124e+00 1.87188172e+00
-4.70330179e-01 3.61280143e-01 2.63141155e-01 -7.21712112e-01
8.13200235e-01 -1.04932979e-01 1.05388984e-01 -7.07995474e-01
-8.50734068e-05 2.10849956e-01 -1.03439651e-01 -3.27426910e-01
5.50346851e-01 1.34080514e-01 2.73357004e-01 1.93200409e-01
-2.22247124e-01 -2.14412540e-01 2.08692551e-02 2.91154265e-01
8.30301344e-01 2.72154659e-01 4.48450707e-02 -2.85606921e-01
6.20266020e-01 -9.00886580e-02 8.96891177e-01 4.40173328e-01
-2.57860217e-02 1.11916387e+00 2.97320664e-01 -4.25612241e-01
-1.03501141e+00 -1.06983650e+00 2.02562567e-02 6.98849678e-01
6.72632456e-01 -5.25967062e-01 -9.85644996e-01 -6.15449250e-01
-2.21025914e-01 5.49344599e-01 -5.05445063e-01 6.17751991e-03
-7.72310495e-01 -7.08378851e-01 1.58414915e-01 4.77087170e-01
9.96645987e-01 -1.18564475e+00 -1.66820049e-01 8.75260532e-02
-5.42334914e-01 -1.25774276e+00 -9.53163743e-01 -2.65024155e-01
-9.53445733e-01 -9.34823871e-01 -1.02629173e+00 -8.11546445e-01
7.65497327e-01 6.58688664e-01 1.02331781e+00 4.83288586e-01
-1.62075490e-01 -1.01756744e-01 -2.30053246e-01 -2.25832552e-01
-3.47106695e-01 4.26131226e-02 -3.50926578e-01 2.06281170e-01
5.26448041e-02 -4.97983187e-01 -8.59887719e-01 5.83569467e-01
-1.14045358e+00 8.68700981e-01 9.20796216e-01 9.02984619e-01
8.14651787e-01 -1.20457700e-02 3.22727352e-01 -9.14244950e-01
1.94535762e-01 -2.04010740e-01 -5.95709443e-01 1.98356658e-01
-5.42979240e-01 -6.18188679e-02 6.25635862e-01 -3.37173522e-01
-1.19950676e+00 -7.70828202e-02 -2.09666103e-01 -6.19008899e-01
-1.84222803e-01 1.24257989e-01 -6.79118693e-01 2.44224607e-03
1.79394662e-01 5.38494468e-01 1.44767731e-01 -7.19757497e-01
2.18219265e-01 4.73199099e-01 6.46955073e-01 -7.00510681e-01
9.64160740e-01 4.69518900e-01 -2.72269398e-01 -5.60414016e-01
-9.29590881e-01 -1.90308332e-01 -3.55960310e-01 -9.94757712e-02
8.68863702e-01 -1.03530419e+00 -2.93584734e-01 9.50777471e-01
-1.11059821e+00 -7.81451702e-01 4.80840281e-02 5.36131077e-02
-4.30701822e-01 6.98331714e-01 -8.84479940e-01 -4.66774464e-01
-5.88509858e-01 -1.52375340e+00 1.16742647e+00 4.21092361e-01
3.07423562e-01 -5.15019655e-01 -4.53299344e-01 7.02488244e-01
4.91774619e-01 1.66374460e-01 7.13523686e-01 9.39406976e-02
-1.14510155e+00 4.29513007e-01 -5.51802337e-01 4.19896990e-01
3.62475455e-01 -9.30026621e-02 -8.92583609e-01 -2.16226771e-01
-1.64624259e-01 4.88675050e-02 1.18256688e+00 2.93384701e-01
1.53169584e+00 -2.82137305e-01 -1.91364408e-01 9.09763634e-01
1.25667357e+00 8.46951157e-02 8.96488190e-01 4.24477518e-01
1.17110741e+00 3.57655704e-01 9.73693430e-01 1.26310512e-01
4.19927269e-01 8.03154051e-01 5.81990778e-01 -4.08828944e-01
-4.92829531e-01 -2.66150534e-01 3.35263699e-01 8.53630543e-01
-1.07685052e-01 -1.23875298e-01 -7.07208514e-01 4.51105446e-01
-2.09112930e+00 -7.55452454e-01 -3.67561162e-01 2.04674268e+00
1.08126807e+00 3.23842615e-01 3.11446656e-03 -7.27835968e-02
7.98386037e-01 3.04114163e-01 -5.86189568e-01 -1.03510603e-01
-1.74853861e-01 1.37541354e-01 3.04048479e-01 5.82262397e-01
-1.15485358e+00 1.25048125e+00 4.78423786e+00 1.25459397e+00
-1.07252169e+00 1.73501119e-01 9.35603142e-01 1.55095205e-01
-4.24063742e-01 -9.36222896e-02 -6.93718731e-01 8.39948416e-01
2.24250346e-01 1.57220602e-01 4.67043817e-01 7.30078518e-01
3.54586333e-01 -2.05325782e-01 -8.00267756e-01 1.16157484e+00
-5.53255342e-03 -1.41910076e+00 4.01802450e-01 -1.03493840e-01
7.47509301e-01 -3.71275276e-01 1.12663217e-01 1.18643776e-01
-1.80586055e-01 -9.85368073e-01 8.66638958e-01 4.62585360e-01
9.00004864e-01 -6.50565982e-01 6.75423086e-01 3.02955449e-01
-1.31014717e+00 8.24114233e-02 -5.18969119e-01 2.01109312e-02
2.59950399e-01 1.04571450e+00 -4.97640997e-01 7.93213308e-01
6.54236495e-01 7.07961857e-01 -7.36209810e-01 1.08021176e+00
-3.26737225e-01 6.08882546e-01 -6.91281185e-02 4.75602180e-01
1.57149613e-01 -4.30048734e-01 4.76466537e-01 1.23830926e+00
2.46969417e-01 -4.22698725e-03 3.29392642e-01 1.05755246e+00
-7.70637468e-02 1.49093885e-02 -2.25854144e-01 1.30538285e-01
3.08136463e-01 1.45312369e+00 -8.47915471e-01 -4.65804756e-01
-3.34760994e-01 1.31006646e+00 5.29209897e-02 3.22844058e-01
-1.03209126e+00 -2.29640543e-01 6.99162185e-01 3.29972178e-01
3.43811661e-01 -2.58384079e-01 -5.48474312e-01 -1.49067998e+00
3.44976038e-01 -1.22200537e+00 1.16576366e-01 -8.62013221e-01
-1.16783929e+00 6.30084395e-01 -4.72612344e-02 -1.32495654e+00
2.43548617e-01 -5.18391848e-01 -7.91904867e-01 8.22753608e-01
-1.52595437e+00 -1.30371177e+00 -7.81498373e-01 4.59984750e-01
1.00370622e+00 3.51405926e-02 2.75668830e-01 5.31408429e-01
-7.22773254e-01 7.19372988e-01 -1.88826829e-01 1.98353082e-01
8.83784831e-01 -1.17739785e+00 7.36844540e-01 1.19249940e+00
4.52033728e-02 4.76129532e-01 6.24720335e-01 -8.11859310e-01
-1.29066384e+00 -1.48949301e+00 3.86258006e-01 -2.28336439e-01
2.23476171e-01 -5.26300490e-01 -1.19932890e+00 6.47872448e-01
4.26926881e-01 -1.57463849e-02 2.46792715e-02 -3.41370851e-01
-2.74029225e-01 -3.01704079e-01 -9.66226816e-01 8.60949993e-01
1.21151400e+00 -6.83983564e-02 -3.88277441e-01 1.99076205e-01
1.01889443e+00 -7.30028570e-01 -5.00937283e-01 5.14170825e-01
3.57884973e-01 -1.19894302e+00 9.31165934e-01 2.08360255e-02
7.96196699e-01 -8.05036604e-01 7.43645132e-02 -1.16597986e+00
-2.62114853e-01 -6.68842375e-01 -2.17494458e-01 1.41924167e+00
7.75569528e-02 -5.82219124e-01 7.68981814e-01 4.52652663e-01
-4.16687220e-01 -9.79635596e-01 -5.18357575e-01 -6.82770729e-01
-5.66396825e-02 -1.80801049e-01 6.87049747e-01 8.90015721e-01
-6.60028219e-01 1.60023078e-01 -6.06369257e-01 1.04440883e-01
7.12506592e-01 5.23988426e-01 1.04935825e+00 -6.54770672e-01
-4.35546190e-01 -3.93286139e-01 -1.03600562e-01 -1.39241588e+00
-2.18698114e-01 -5.94873846e-01 1.73029795e-01 -1.64811575e+00
2.96361655e-01 -5.57668149e-01 1.25610000e-02 5.51387370e-01
-6.28443122e-01 3.10293972e-01 4.01370555e-01 2.43813738e-01
-5.35419106e-01 7.44120419e-01 1.76156497e+00 -3.36813748e-01
1.69766471e-02 -8.27497467e-02 -6.18478656e-01 8.13681543e-01
9.93871033e-01 -2.35541269e-01 -3.94068897e-01 -7.04755962e-01
-2.76103437e-01 -1.03795417e-01 5.25996506e-01 -1.00379610e+00
-3.27354819e-02 -4.65799034e-01 4.73324597e-01 -7.03112602e-01
4.14841294e-01 -6.17367446e-01 2.74211138e-01 3.89025569e-01
2.89847003e-03 6.23296909e-02 2.32877776e-01 5.19829750e-01
-2.21393585e-01 -8.41141399e-03 9.11983311e-01 -2.55758375e-01
-8.57390583e-01 5.99958539e-01 2.30983421e-02 -2.41959700e-04
8.33324730e-01 -1.77771989e-02 -4.02322769e-01 -3.04059654e-01
-3.86689812e-01 2.35141948e-01 9.03985262e-01 3.90392542e-01
8.24911177e-01 -1.33297431e+00 -6.41459227e-01 2.48966455e-01
-5.08380532e-02 5.80197752e-01 4.63595361e-01 9.45798755e-01
-7.67075121e-01 -2.80249361e-02 -2.07960084e-01 -5.90038717e-01
-1.16996396e+00 6.68376744e-01 2.22856417e-01 -1.85941726e-01
-9.94732082e-01 7.21842229e-01 9.07098413e-01 -1.72028318e-01
2.03683183e-01 -5.20397484e-01 2.76643664e-01 -5.40544152e-01
7.89205134e-01 1.23476639e-01 -4.46626954e-02 -4.32582706e-01
-1.11910380e-01 5.57521999e-01 -3.52484494e-01 1.44951329e-01
1.06611848e+00 -3.09541613e-01 -2.50938207e-01 2.27496371e-01
6.40461445e-01 -2.69811023e-02 -1.44728506e+00 -2.56011367e-01
-2.32899055e-01 -8.69924307e-01 -1.49834126e-01 -7.20031559e-01
-1.41796553e+00 1.06522703e+00 4.71575916e-01 -1.28287435e-01
1.35517824e+00 -2.11080328e-01 1.30907130e+00 -2.99681141e-03
3.66122901e-01 -7.38348544e-01 1.82489574e-01 2.21438900e-01
1.13639522e+00 -1.35578978e+00 2.01921421e-03 -8.84660184e-01
-6.44692183e-01 9.47799027e-01 1.11952662e+00 -1.91283524e-01
2.12978065e-01 2.56108522e-01 2.53221720e-01 7.14675188e-02
-4.92098182e-01 -3.14239301e-02 3.18788290e-01 4.66043919e-01
2.86523223e-01 7.72619843e-02 -3.63156229e-01 6.31670237e-01
-5.38794026e-02 -6.97216615e-02 4.56743181e-01 6.63285077e-01
-3.58048201e-01 -1.21783471e+00 -5.22200882e-01 3.90321404e-01
-3.70923221e-01 -3.17539364e-01 -1.47203952e-01 5.68718433e-01
2.51640111e-01 8.09266508e-01 -1.05856434e-01 -3.09264213e-01
1.66302949e-01 -1.81622848e-01 4.38750893e-01 -5.51169813e-01
-3.65157694e-01 3.47560078e-01 -1.65272996e-01 -8.02496076e-01
-2.41247967e-01 -3.86566997e-01 -1.15757382e+00 -4.38039780e-01
-1.97154790e-01 -9.41026732e-02 2.92404711e-01 7.26722121e-01
4.62181866e-01 6.47829473e-01 2.97459483e-01 -9.40860629e-01
-1.48586184e-01 -9.43706632e-01 -3.54996055e-01 4.65443790e-01
3.22718889e-01 -6.44796610e-01 -1.01098724e-01 2.95752883e-01] | [10.642117500305176, -0.8910134434700012] |
b5ad31c1-429c-4713-84ac-02edc0067011 | dynamic-graph-cnn-for-learning-on-point | 1801.07829 | null | https://arxiv.org/abs/1801.07829v2 | https://arxiv.org/pdf/1801.07829v2.pdf | Dynamic Graph CNN for Learning on Point Clouds | Point clouds provide a flexible geometric representation suitable for countless applications in computer graphics; they also comprise the raw output of most 3D data acquisition devices. While hand-designed features on point clouds have long been proposed in graphics and vision, however, the recent overwhelming success of convolutional neural networks (CNNs) for image analysis suggests the value of adapting insight from CNN to the point cloud world. Point clouds inherently lack topological information so designing a model to recover topology can enrich the representation power of point clouds. To this end, we propose a new neural network module dubbed EdgeConv suitable for CNN-based high-level tasks on point clouds including classification and segmentation. EdgeConv acts on graphs dynamically computed in each layer of the network. It is differentiable and can be plugged into existing architectures. Compared to existing modules operating in extrinsic space or treating each point independently, EdgeConv has several appealing properties: It incorporates local neighborhood information; it can be stacked applied to learn global shape properties; and in multi-layer systems affinity in feature space captures semantic characteristics over potentially long distances in the original embedding. We show the performance of our model on standard benchmarks including ModelNet40, ShapeNetPart, and S3DIS. | ['Yue Wang', 'Yongbin Sun', 'Michael M. Bronstein', 'Ziwei Liu', 'Sanjay E. Sarma', 'Justin M. Solomon'] | 2018-01-24 | null | null | null | null | ['3d-part-segmentation', 'few-shot-3d-point-cloud-classification'] | ['computer-vision', 'computer-vision'] | [-3.93267006e-01 8.84310305e-02 -9.59487781e-02 -5.18847704e-01
-8.47654268e-02 -6.58973873e-01 7.62207568e-01 3.10281575e-01
-2.64371544e-01 -1.56866223e-01 -1.46481007e-01 -5.17370105e-01
-1.20245606e-01 -1.12343955e+00 -9.11138892e-01 -1.27783656e-01
-2.58034080e-01 7.86086738e-01 4.44145858e-01 -3.62463087e-01
3.60534430e-01 1.49944675e+00 -1.57467687e+00 1.68409139e-01
6.20615840e-01 1.12356138e+00 1.00299157e-01 4.14394170e-01
-5.71053326e-01 9.37960576e-03 -1.62822187e-01 -2.23792449e-01
4.91907716e-01 4.02394772e-01 -5.17748594e-01 -1.54075041e-01
8.47116470e-01 -1.36179656e-01 -3.26822937e-01 8.49228859e-01
3.61885041e-01 1.37667479e-02 5.82607329e-01 -1.24346638e+00
-9.12794173e-01 1.17131777e-01 -3.22100788e-01 -3.79386246e-02
7.55371898e-02 1.62928000e-01 1.12415838e+00 -1.30958045e+00
7.69828022e-01 1.49620521e+00 1.09394634e+00 4.63727295e-01
-1.26888096e+00 -3.92089933e-01 2.37489209e-01 -5.69863729e-02
-1.11366856e+00 3.18541564e-02 1.08978069e+00 -4.76488382e-01
1.14492166e+00 3.73495966e-01 1.02489841e+00 7.40184724e-01
1.20821930e-01 6.28096461e-01 5.82704425e-01 1.25626534e-01
2.83600897e-01 -1.68461367e-01 1.39856413e-01 8.39449227e-01
2.49384895e-01 1.89103577e-02 -3.89731169e-01 -2.67750293e-01
1.43431640e+00 4.77382153e-01 -1.32696941e-01 -9.62127090e-01
-1.17243564e+00 8.60899866e-01 1.34260094e+00 8.68108422e-02
-7.96673894e-02 7.13101923e-01 2.87690371e-01 3.22953820e-01
5.96073389e-01 5.80373406e-01 -5.15296161e-01 5.47076538e-02
-7.25903690e-01 4.24667239e-01 5.87139308e-01 1.11525285e+00
1.00949383e+00 1.05942739e-03 2.35068098e-01 5.42983711e-01
3.11153054e-01 4.93299007e-01 6.95834085e-02 -1.01127481e+00
2.87188888e-01 1.17209101e+00 -1.68067500e-01 -1.33702219e+00
-6.65064216e-01 -4.73126978e-01 -8.20595086e-01 7.03899622e-01
4.15990613e-02 4.38417584e-01 -1.12670577e+00 1.15182817e+00
3.60625833e-01 3.47045332e-01 -5.59554398e-01 9.96683359e-01
1.12262559e+00 4.38351661e-01 -2.05438584e-01 9.03674424e-01
1.11357355e+00 -6.57453895e-01 1.00689173e-01 -9.54140127e-02
5.93815744e-01 -4.34901625e-01 1.14164674e+00 1.43210277e-01
-1.07569766e+00 -5.77262342e-01 -1.03681064e+00 -6.68243885e-01
-7.95808017e-01 -1.12904333e-01 1.04923439e+00 3.55414510e-01
-1.51927722e+00 9.41010177e-01 -1.08914256e+00 -3.17268848e-01
9.23514187e-01 6.64238095e-01 -5.10876894e-01 -5.85712492e-02
-3.95820022e-01 5.74429274e-01 1.23872973e-01 9.81509015e-02
-2.76911110e-01 -9.80847538e-01 -9.26505685e-01 2.87687838e-01
5.22830896e-03 -1.17353332e+00 8.55582535e-01 -5.32803178e-01
-1.40443230e+00 8.86220813e-01 8.51206109e-03 -1.99167550e-01
4.71805334e-01 -1.07600704e-01 4.72453646e-02 1.34956032e-01
-4.80958484e-02 9.29056287e-01 8.01418364e-01 -1.20479441e+00
-3.42737228e-01 -6.78790510e-01 1.59695432e-01 6.28307909e-02
3.33814969e-04 -4.93421078e-01 -6.42992020e-01 -4.15688157e-01
7.63564646e-01 -9.94332552e-01 -3.73155475e-01 8.09633315e-01
-2.89501160e-01 -4.84995008e-01 1.29615879e+00 5.65565005e-02
4.79723215e-01 -2.17599845e+00 -3.76259647e-02 4.13766623e-01
6.65303349e-01 2.83813536e-01 -2.46185139e-01 2.96325415e-01
-2.03090429e-01 3.38789523e-01 -2.48987257e-01 -4.32252526e-01
1.21254303e-01 2.29938373e-01 -1.55631617e-01 5.10563374e-01
6.07880294e-01 1.50356424e+00 -9.41588461e-01 -9.38843191e-02
6.10511959e-01 7.85162032e-01 -8.70666862e-01 -1.47966757e-01
-3.70807052e-01 2.24990532e-01 -5.86454690e-01 7.18345881e-01
8.93474877e-01 -5.63441455e-01 -4.03990239e-01 -2.69577771e-01
-1.60609055e-02 2.07991540e-01 -8.26393425e-01 2.08254504e+00
-4.90195513e-01 6.87134683e-01 -7.54601462e-03 -7.63324678e-01
1.09795892e+00 -1.57197967e-01 5.62559903e-01 -4.05024111e-01
1.07057780e-01 1.93827927e-01 -2.98861891e-01 4.26606871e-02
5.90721726e-01 1.55121237e-01 6.46302029e-02 1.59546450e-01
3.07457000e-02 -6.61815345e-01 -5.56130588e-01 1.24686234e-01
1.09585476e+00 4.91652004e-02 -1.99028984e-01 -2.84758836e-01
2.20583588e-01 1.30261436e-01 1.28800884e-01 5.39239466e-01
1.59899786e-01 8.72436166e-01 2.70123124e-01 -9.53185439e-01
-1.21614623e+00 -1.36432958e+00 -3.70303333e-01 8.19105446e-01
3.31159770e-01 -4.90741134e-01 -1.51162177e-01 -6.46744490e-01
6.18082106e-01 3.29349786e-01 -5.89787662e-01 -9.66208726e-02
-6.99004650e-01 -1.28521686e-02 2.67092943e-01 8.07195008e-01
3.56953561e-01 -8.72588634e-01 -5.02478480e-01 1.17130667e-01
6.68307781e-01 -1.00088322e+00 -2.92140216e-01 1.75301582e-01
-1.33982253e+00 -1.08857560e+00 -3.62767398e-01 -6.66861773e-01
7.37226605e-01 5.28779984e-01 1.42607534e+00 3.70632321e-01
-2.09018007e-01 6.52941525e-01 -1.14720874e-01 -5.17537296e-01
7.65635669e-02 3.22457284e-01 -1.89368054e-01 -2.91304648e-01
6.21881545e-01 -9.42678928e-01 -7.99247324e-01 1.56150356e-01
-8.72796297e-01 3.77646312e-02 3.60766441e-01 6.49502993e-01
7.77953506e-01 -4.71599847e-01 1.24859080e-01 -8.51053298e-01
4.68942761e-01 -4.05630291e-01 -6.99176490e-01 -2.25944370e-01
-1.98225766e-01 -2.73355991e-02 5.72914183e-01 -6.40576780e-02
-2.99845427e-01 1.39356434e-01 -3.80976468e-01 -9.44917738e-01
-2.10406631e-01 3.90475750e-01 -1.27183929e-01 -6.55031145e-01
4.06557441e-01 -1.69619054e-01 1.29609838e-01 -7.02327728e-01
6.51622474e-01 1.08371057e-01 5.34231126e-01 -4.27164346e-01
9.50879037e-01 9.03261900e-01 5.22194743e-01 -1.04572332e+00
-3.44230801e-01 -4.81481880e-01 -9.63863015e-01 4.20561135e-02
8.27365935e-01 -7.33453572e-01 -1.05189323e+00 2.90768564e-01
-1.40067506e+00 -1.73899412e-01 -5.03827929e-01 1.25788406e-01
-5.93093336e-01 1.12116128e-01 -2.82968611e-01 -3.09974283e-01
-2.33828083e-01 -1.09097111e+00 1.61121094e+00 3.71588357e-02
4.35573086e-02 -1.45713866e+00 -1.73270315e-01 -2.37275079e-01
3.77396256e-01 5.68278313e-01 1.05519700e+00 -4.05141532e-01
-1.12033641e+00 -3.72155815e-01 -5.48120618e-01 1.04082443e-01
4.79496084e-02 5.90650812e-02 -9.96434748e-01 -3.48142713e-01
-1.59619167e-01 4.43068892e-02 1.06342125e+00 3.42244238e-01
1.71602619e+00 4.43702489e-02 -5.07893384e-01 1.28784299e+00
1.56741714e+00 -3.82409573e-01 4.33700502e-01 2.55970865e-01
1.28557825e+00 2.08282873e-01 -1.21060014e-01 9.22610611e-02
4.64031816e-01 6.57349885e-01 1.02242649e+00 -4.66946989e-01
-1.38600960e-01 -4.21413064e-01 -2.21950665e-01 9.63190258e-01
-1.45645693e-01 7.76558816e-02 -1.14018190e+00 4.29278105e-01
-1.78677905e+00 -4.53404605e-01 -3.69650781e-01 1.91939604e+00
1.45883411e-01 2.32155606e-01 -1.56229004e-01 -1.63459763e-01
2.83918440e-01 1.89926639e-01 -8.27249229e-01 -5.28803408e-01
-1.90814614e-01 5.89197993e-01 7.29987919e-01 1.93578273e-01
-1.06714702e+00 1.02743053e+00 6.17057133e+00 5.55549383e-01
-1.48057175e+00 -1.07968196e-01 4.23322231e-01 -2.21190788e-02
-6.88226521e-01 -1.08748659e-01 -5.01364410e-01 1.23683617e-01
5.34775496e-01 2.73407269e-02 4.44416292e-02 1.10877097e+00
-8.54912400e-02 4.65921760e-01 -1.51056182e+00 1.28746939e+00
-2.63544321e-01 -1.91997671e+00 4.88977969e-01 2.92565286e-01
5.02356410e-01 7.08851099e-01 3.14683855e-01 -3.49583030e-02
3.44931424e-01 -1.33292329e+00 6.71233594e-01 5.19202828e-01
1.06226254e+00 -7.37381935e-01 4.10973519e-01 2.43920177e-01
-1.27700472e+00 1.65341780e-01 -8.06727171e-01 1.00068776e-02
-4.63906415e-02 4.99121040e-01 -9.02225018e-01 5.06365120e-01
6.86969638e-01 1.14756477e+00 -7.55924284e-01 1.22586095e+00
8.43936354e-02 2.11981907e-01 -6.55064702e-01 3.27331312e-02
6.44968092e-01 -4.00591582e-01 6.90552652e-01 1.03677154e+00
3.45274448e-01 -1.48834690e-01 1.79685608e-01 1.28167319e+00
-2.51714319e-01 -1.13793023e-01 -1.13912582e+00 1.21957801e-01
4.32867706e-01 1.15211475e+00 -8.76078963e-01 -6.65130615e-02
-4.52224582e-01 7.64349937e-01 6.07140064e-01 2.43301436e-01
-3.36940229e-01 -4.14467633e-01 1.05921090e+00 4.26643610e-01
5.01859725e-01 -8.91405225e-01 -7.27882683e-01 -1.01450133e+00
5.72719611e-02 -1.91078588e-01 -7.66175613e-02 -8.30541790e-01
-1.47856748e+00 6.65103078e-01 -3.14342797e-01 -1.29031098e+00
1.43627167e-01 -1.06769514e+00 -7.83064604e-01 9.55611229e-01
-1.59394622e+00 -1.37782025e+00 -5.50544918e-01 6.20491087e-01
4.26687807e-01 -5.13754378e-04 7.22099543e-01 1.72747206e-02
5.57447411e-02 3.64568353e-01 6.56335652e-02 1.89882249e-01
5.36985360e-02 -1.47007549e+00 1.40907741e+00 2.78415143e-01
3.96588951e-01 8.06437790e-01 5.61918281e-02 -5.92465401e-01
-1.89003980e+00 -1.31061935e+00 4.12546962e-01 -8.17100823e-01
5.27748406e-01 -7.48362958e-01 -1.17165041e+00 6.01524591e-01
-3.81722689e-01 4.10799354e-01 2.21587211e-01 1.17272474e-01
-6.44638240e-01 1.02686293e-01 -1.00303209e+00 6.00535989e-01
1.45311785e+00 -6.89061463e-01 -4.11342263e-01 2.33079985e-01
1.00907779e+00 -6.45214677e-01 -8.46082449e-01 3.92675996e-01
2.72747844e-01 -9.95095670e-01 1.37978649e+00 -8.22936118e-01
1.81715190e-01 -3.12857270e-01 -1.37212696e-02 -1.36776543e+00
-6.96123540e-01 -4.07760680e-01 -5.21575063e-02 5.08169115e-01
2.00112626e-01 -7.78263688e-01 1.26020503e+00 5.94690621e-01
-5.40802360e-01 -9.13852334e-01 -1.08094287e+00 -7.18532383e-01
3.55054110e-01 -6.59983158e-01 1.06647217e+00 1.02113855e+00
-5.59708178e-01 2.11786434e-01 4.15015996e-01 2.76640356e-01
6.24039173e-01 2.20127732e-01 9.16262150e-01 -1.90609205e+00
2.20195964e-01 -8.92218173e-01 -1.02973294e+00 -1.46993256e+00
1.61678359e-01 -1.38706195e+00 -4.91580844e-01 -1.68643725e+00
-4.60547388e-01 -1.00273895e+00 -5.60750887e-02 4.61959660e-01
1.67747900e-01 4.07711625e-01 2.22211987e-01 2.89076656e-01
-2.62705982e-01 7.36366034e-01 1.41133296e+00 -2.28798792e-01
-1.67401642e-01 1.22373039e-02 -4.23928738e-01 8.45127165e-01
5.61360359e-01 -1.05653882e-01 -2.54011720e-01 -9.38486218e-01
5.41595221e-01 -4.11646366e-01 8.28318954e-01 -1.02106082e+00
3.42432261e-01 5.11461496e-02 4.98888463e-01 -9.52687025e-01
5.96280754e-01 -1.22624123e+00 1.22454084e-01 5.81881329e-02
1.52655274e-01 2.62525231e-01 3.15561831e-01 6.37366176e-01
-1.31646067e-01 1.20042659e-01 5.75618088e-01 -3.46019536e-01
-6.79417193e-01 9.13854420e-01 3.42411935e-01 -2.81682134e-01
7.00156510e-01 -7.38888621e-01 -2.03627944e-01 -7.27961063e-02
-5.80545545e-01 2.31182292e-01 1.11174989e+00 6.19110346e-01
1.03624332e+00 -1.62701488e+00 -3.74123216e-01 4.50519264e-01
1.87124759e-01 7.21139252e-01 -7.02611506e-02 4.23817694e-01
-1.08470809e+00 4.22052145e-01 -2.58473903e-01 -1.23439717e+00
-8.39852154e-01 5.15891016e-01 3.45361501e-01 3.69102806e-01
-1.21847522e+00 8.87692094e-01 4.48257416e-01 -7.11766303e-01
7.24041909e-02 -8.71872842e-01 4.28563394e-02 -3.57939303e-01
7.04867840e-02 2.28577062e-01 4.18117076e-01 -6.55393660e-01
-3.45450103e-01 8.93070221e-01 1.72682762e-01 3.14625114e-01
1.75204170e+00 2.43966058e-01 -3.06040853e-01 6.39313102e-01
1.49709749e+00 -2.89204895e-01 -1.20799875e+00 -2.95278549e-01
1.46307290e-01 -6.54198110e-01 2.09075913e-01 -2.43916839e-01
-1.17798865e+00 1.21498024e+00 3.75882566e-01 3.86977911e-01
5.28891087e-01 2.17433214e-01 7.60026217e-01 5.95743716e-01
6.35095239e-01 -5.75639367e-01 4.20420915e-02 7.85811841e-01
1.02457583e+00 -1.11309445e+00 -5.53598441e-02 -6.50872707e-01
-9.19204429e-02 1.30960548e+00 4.14485544e-01 -7.03528881e-01
1.09528601e+00 6.01384789e-03 -9.46684480e-02 -8.55324745e-01
-5.16593158e-01 2.54591666e-02 6.88442588e-01 7.88239777e-01
1.54646307e-01 1.08853407e-01 3.97685260e-01 9.87477154e-02
-5.01918137e-01 -3.23884845e-01 2.88996309e-01 8.93055797e-01
-4.61296916e-01 -9.20399666e-01 -1.89143285e-01 6.85080051e-01
6.38563782e-02 5.76071069e-02 -5.89301109e-01 7.82403231e-01
2.63492931e-02 1.62435681e-01 5.97799659e-01 -2.67741382e-01
5.57033777e-01 -2.26326168e-01 2.37915337e-01 -7.95051634e-01
-6.54126227e-01 -3.51569980e-01 -3.12899530e-01 -9.55480039e-01
-1.25667259e-01 -3.68538737e-01 -1.19910586e+00 -4.48514581e-01
-1.41785562e-01 -9.66321379e-02 1.14501977e+00 4.77303684e-01
8.20592225e-01 3.81631196e-01 5.04669785e-01 -1.51558030e+00
-1.73845634e-01 -5.60105562e-01 -5.91860771e-01 5.88360608e-01
4.09537792e-01 -7.37063885e-01 -2.22162679e-01 -3.61226916e-01] | [7.921504020690918, -3.717244863510132] |
b30ab5a4-dc71-49c7-9b05-712ae8648c22 | video-prediction-by-efficient-transformers | 2212.06026 | null | https://arxiv.org/abs/2212.06026v1 | https://arxiv.org/pdf/2212.06026v1.pdf | Video Prediction by Efficient Transformers | Video prediction is a challenging computer vision task that has a wide range of applications. In this work, we present a new family of Transformer-based models for video prediction. Firstly, an efficient local spatial-temporal separation attention mechanism is proposed to reduce the complexity of standard Transformers. Then, a full autoregressive model, a partial autoregressive model and a non-autoregressive model are developed based on the new efficient Transformer. The partial autoregressive model has a similar performance with the full autoregressive model but a faster inference speed. The non-autoregressive model not only achieves a faster inference speed but also mitigates the quality degradation problem of the autoregressive counterparts, but it requires additional parameters and loss function for learning. Given the same attention mechanism, we conducted a comprehensive study to compare the proposed three video prediction variants. Experiments show that the proposed video prediction models are competitive with more complex state-of-the-art convolutional-LSTM based models. The source code is available at https://github.com/XiYe20/VPTR. | ['Guillaume-Alexandre Bilodeau', 'Xi Ye'] | 2022-12-12 | null | null | null | null | ['video-prediction'] | ['computer-vision'] | [ 7.44571164e-02 -1.05787523e-01 -2.04879731e-01 -2.22001076e-01
-5.87012231e-01 1.21331476e-01 5.82626283e-01 -4.72259521e-01
-2.46833339e-01 3.94471079e-01 3.22646230e-01 -2.40993947e-01
1.59699455e-01 -5.42999566e-01 -9.29452419e-01 -7.50518501e-01
2.18447834e-01 5.46406284e-02 7.32003868e-01 5.52857481e-02
3.00427545e-02 9.77731943e-02 -1.45082414e+00 3.73570651e-01
9.61458445e-01 1.44306386e+00 6.48328006e-01 4.78160709e-01
-5.44750728e-02 1.34891844e+00 4.10983600e-02 -4.36286777e-01
1.24552019e-01 -2.61999756e-01 -4.33689743e-01 1.86811830e-03
3.52194577e-01 -5.63156188e-01 -7.64987528e-01 9.41798389e-01
3.05307686e-01 1.49745211e-01 3.86038780e-01 -1.25904930e+00
-7.52562761e-01 5.18509626e-01 -4.84660119e-01 2.62821883e-01
2.74917688e-02 -1.15785308e-01 8.40279400e-01 -1.02251959e+00
1.17046587e-01 1.21047068e+00 7.02643275e-01 4.01116729e-01
-7.68353820e-01 -9.38042104e-01 5.91086864e-01 7.69538760e-01
-1.27680218e+00 -5.93362987e-01 7.24178433e-01 -4.03648764e-01
1.05945516e+00 -3.30680497e-02 7.05956876e-01 1.14386714e+00
4.28674966e-01 1.05927205e+00 8.17072928e-01 -2.16314197e-01
-1.13007724e-01 5.10374531e-02 1.53804326e-03 9.02190864e-01
-3.90509188e-01 -1.47401746e-02 -3.36290240e-01 7.46744797e-02
7.98326492e-01 4.42108303e-01 -5.75177014e-01 -4.13107485e-01
-8.61032903e-01 6.66962147e-01 4.94256526e-01 2.24979132e-01
-5.43137729e-01 2.90569842e-01 5.39198041e-01 1.44732803e-01
6.64265633e-01 -1.89606771e-01 -3.86479080e-01 -2.19760999e-01
-9.15374756e-01 -7.78595433e-02 4.07566220e-01 1.04037464e+00
4.83590424e-01 6.12823606e-01 -2.55730301e-01 9.73351300e-01
6.02720439e-01 4.05604064e-01 6.56539440e-01 -8.66001844e-01
5.66183746e-01 2.45925471e-01 -2.68674102e-02 -1.03674972e+00
-1.26196504e-01 -4.36446995e-01 -1.05684221e+00 -1.78652450e-01
-5.21060340e-02 2.39552781e-02 -8.74164104e-01 1.51336312e+00
1.34504372e-02 8.24509799e-01 -1.66994363e-01 8.18524003e-01
8.84236097e-01 1.12765956e+00 1.68585524e-01 -4.15393502e-01
1.08806419e+00 -1.62760723e+00 -7.91075647e-01 -1.91336185e-01
3.34503621e-01 -7.89214492e-01 9.33661520e-01 4.47326362e-01
-1.17012084e+00 -8.18590641e-01 -9.05373812e-01 -1.46235615e-01
1.95842087e-01 4.71620262e-01 4.53256190e-01 3.27470720e-01
-1.29369068e+00 5.17034888e-01 -1.09210753e+00 -2.72618890e-01
2.84769595e-01 2.94007897e-01 -1.82149380e-01 -1.14407830e-01
-1.13710690e+00 7.40591228e-01 1.85650840e-01 3.17655623e-01
-9.44364905e-01 -5.25196493e-01 -1.01906919e+00 5.23062050e-01
3.69836211e-01 -5.49550891e-01 1.42387557e+00 -1.30567992e+00
-1.83699012e+00 3.48308980e-01 -5.85273802e-01 -6.30402744e-01
3.41944456e-01 -5.94258487e-01 -3.40460569e-01 9.06794518e-02
-2.38454625e-01 5.08349657e-01 1.08887792e+00 -8.31642091e-01
-6.30127847e-01 -8.74548554e-02 -7.06290156e-02 2.22226530e-01
-5.77944100e-01 4.67926487e-02 -8.46880674e-01 -7.93747187e-01
-1.73001096e-01 -9.12967086e-01 -1.26977071e-01 -7.53765255e-02
1.42773660e-02 -1.86187714e-01 9.86960411e-01 -8.91492009e-01
1.41112387e+00 -2.13387418e+00 2.60330528e-01 -2.03644291e-01
1.79307640e-01 5.68006098e-01 -2.40642518e-01 2.93579161e-01
-1.75575897e-01 -7.45574608e-02 1.07840359e-01 -4.67779547e-01
-1.23811193e-01 -3.51679586e-02 -5.42796493e-01 2.05135226e-01
8.31953958e-02 7.07633197e-01 -5.78801513e-01 -5.33292592e-01
4.75475788e-01 7.92257071e-01 -5.98277092e-01 3.25438678e-01
-9.79858786e-02 2.89900362e-01 -3.92587632e-01 4.41428274e-01
6.65863752e-01 -3.32616866e-01 2.10088253e-01 -4.50166821e-01
-2.22860858e-01 3.24465901e-01 -7.67826080e-01 1.29951942e+00
-5.02445757e-01 6.32723629e-01 -1.44617051e-01 -1.23615754e+00
9.69633877e-01 5.39631188e-01 3.97942185e-01 -7.05688059e-01
1.12899177e-01 9.03664455e-02 -2.84089725e-02 -5.19458711e-01
4.71226752e-01 -1.40013188e-01 4.92599547e-01 -5.68235852e-02
1.77828655e-01 4.65055197e-01 3.70985791e-02 -2.54046428e-03
8.89297247e-01 5.44149041e-01 2.13541180e-01 -1.27243444e-01
8.00726891e-01 -4.16811645e-01 1.06789172e+00 4.02903616e-01
-2.55648762e-01 4.13195670e-01 3.77076328e-01 -5.32908380e-01
-8.76161218e-01 -9.20712948e-01 3.03094089e-01 1.05370581e+00
2.33790621e-01 -6.01283371e-01 -4.75179285e-01 -4.25348192e-01
-4.43457305e-01 6.35423899e-01 -3.85009915e-01 -1.85552552e-01
-5.90179205e-01 -3.76645386e-01 1.27459243e-01 8.32807720e-01
8.52300465e-01 -9.18823659e-01 -2.88932204e-01 1.15851440e-01
-3.92370671e-01 -1.25582457e+00 -6.20160460e-01 -2.97474205e-01
-9.66162562e-01 -6.89005077e-01 -8.35899293e-01 -1.00274837e+00
2.50196904e-01 4.35434103e-01 8.97676945e-01 9.64316055e-02
4.27155524e-01 4.56944466e-01 -5.87349117e-01 -2.05890432e-01
-1.71025336e-01 -2.18516648e-01 -6.76642656e-02 2.41931051e-01
3.42598647e-01 -5.36359131e-01 -6.05052590e-01 3.97783428e-01
-6.17725015e-01 3.88181001e-01 6.70376658e-01 1.01178849e+00
7.36479878e-01 -5.50541319e-02 3.91924053e-01 -6.27667904e-01
1.60582364e-01 -4.27029282e-01 -7.04561651e-01 2.31995121e-01
-3.73763889e-01 -7.87813682e-03 1.03397739e+00 -5.26335061e-01
-1.40418220e+00 2.47812673e-01 -3.03291827e-01 -1.11082780e+00
1.89188033e-01 5.19396245e-01 -1.25775486e-01 3.33974808e-02
-1.15821123e-01 5.81015110e-01 -1.32131845e-01 -6.16151869e-01
-1.89027227e-02 4.66937900e-01 2.50910372e-01 -1.32061645e-01
5.15262425e-01 2.66651839e-01 -3.00101727e-01 -8.69591177e-01
-6.39508963e-01 -3.41450453e-01 -3.45985740e-01 -3.12559843e-01
9.32435155e-01 -1.24533331e+00 -4.58072871e-01 8.01547945e-01
-1.21143687e+00 -4.83709931e-01 2.00971842e-01 7.67069519e-01
-7.05285728e-01 5.67188323e-01 -9.69547570e-01 -8.01025629e-01
-4.93335515e-01 -1.40957272e+00 8.35212052e-01 1.05945103e-01
2.35085294e-01 -9.57369506e-01 -1.56244352e-01 3.15080613e-01
4.34216797e-01 -2.99167424e-01 6.36413991e-01 -4.59882706e-01
-8.13105524e-01 -8.49617936e-04 -2.87159085e-01 5.20109296e-01
-3.67853157e-02 1.56597689e-01 -8.63823771e-01 -2.96770394e-01
1.03397049e-01 -1.02049120e-01 1.11607254e+00 7.27068365e-01
1.31810498e+00 -3.31752926e-01 -3.45493108e-01 7.86548257e-01
1.31765831e+00 5.57880402e-01 9.78044391e-01 2.94599205e-01
7.44449615e-01 1.45716682e-01 7.08005965e-01 3.86466801e-01
5.77654541e-01 8.08935821e-01 3.87701094e-01 -1.24620460e-02
1.38100296e-01 -4.04822499e-01 6.69252038e-01 1.26912737e+00
-3.19916368e-01 -3.17599863e-01 -7.65122831e-01 3.51556748e-01
-2.19167662e+00 -1.21980834e+00 -1.42435834e-01 2.17312932e+00
2.95202076e-01 1.06858082e-01 2.96101905e-02 -8.37930217e-02
7.35960543e-01 3.52407992e-01 -3.96189451e-01 -4.38257545e-01
2.79734910e-01 8.99631605e-02 3.43297243e-01 4.02354002e-01
-1.38464296e+00 1.02520657e+00 5.66633654e+00 9.93463397e-01
-1.23204625e+00 1.57964692e-01 6.71431065e-01 5.17258719e-02
5.77140637e-02 -1.45505309e-01 -7.43091822e-01 7.36251593e-01
9.84872162e-01 -1.44410819e-01 2.75265247e-01 1.01925886e+00
5.08713782e-01 1.59311265e-01 -8.58381450e-01 1.16611600e+00
2.27508932e-01 -1.29788792e+00 1.82011664e-01 -2.08491102e-01
3.80979151e-01 2.27161631e-01 1.37922734e-01 4.64198351e-01
2.39487588e-02 -8.23733509e-01 7.66451359e-01 7.17972159e-01
4.72469985e-01 -6.32743537e-01 8.11383426e-01 2.28474304e-01
-1.67570305e+00 -4.36567932e-01 -5.90849280e-01 -1.73052520e-01
3.11337054e-01 3.14157218e-01 -1.21274248e-01 4.56791997e-01
9.90231276e-01 1.18662107e+00 -2.67885387e-01 1.18024290e+00
-1.90942243e-01 7.66113281e-01 -2.18457021e-02 2.90427893e-01
2.46039391e-01 -3.15770358e-01 4.24420089e-01 1.10815775e+00
7.46784627e-01 1.69990152e-01 8.77797678e-02 3.92652571e-01
-3.80575098e-02 1.14447415e-01 -4.19907689e-01 1.05297141e-01
3.55534554e-01 1.10959375e+00 -3.22332233e-01 -6.17050231e-01
-8.76919925e-01 1.05650532e+00 3.17670047e-01 4.40937489e-01
-1.28221810e+00 -1.09769695e-01 4.99093026e-01 3.70820686e-02
8.14377904e-01 -3.83964404e-02 3.27185333e-01 -1.43942988e+00
8.81047472e-02 -8.00693393e-01 2.81010181e-01 -1.05820990e+00
-9.90605295e-01 8.56325030e-01 -1.57425359e-01 -1.63895774e+00
-3.55119586e-01 -5.48114538e-01 -8.19252193e-01 6.16426826e-01
-1.62758362e+00 -1.11590946e+00 -5.01244366e-01 5.87923050e-01
9.84030247e-01 -2.21875429e-01 5.82742214e-01 5.66974282e-01
-7.79101551e-01 5.99644840e-01 2.04591870e-01 -6.85270652e-02
6.28719151e-01 -7.66964793e-01 1.45103574e-01 9.95211124e-01
-1.61133349e-01 4.21584010e-01 5.97681582e-01 -4.13635403e-01
-1.22404838e+00 -1.31203187e+00 7.79363036e-01 1.78589463e-01
7.14073062e-01 -6.60931244e-02 -1.13311315e+00 1.07190228e+00
3.70352656e-01 -1.21020183e-01 4.62157279e-01 -2.70913124e-01
-3.16443443e-01 -3.85097831e-01 -6.61455095e-01 5.02675116e-01
7.53574193e-01 -3.31878453e-01 -2.99357712e-01 1.54121414e-01
8.22037995e-01 -2.41904527e-01 -8.77230465e-01 4.86477017e-01
7.33911037e-01 -1.23989940e+00 8.75286460e-01 -2.17776760e-01
7.44627178e-01 -1.78651601e-01 -1.49574792e-02 -1.07954574e+00
-7.34516919e-01 -4.25448895e-01 -4.09607440e-01 1.15837133e+00
2.21630782e-01 -7.17075348e-01 7.14544237e-01 4.92850006e-01
-4.89403307e-01 -1.11536825e+00 -7.89812863e-01 -8.54302049e-01
-1.47701383e-01 -3.68577242e-01 2.08074167e-01 5.79637170e-01
-1.16649538e-01 3.29518318e-01 -9.40207541e-01 1.20219313e-01
4.31897819e-01 7.02142343e-02 5.44290960e-01 -7.80508876e-01
-5.03592253e-01 -1.95551783e-01 -4.55813169e-01 -1.68742418e+00
2.30938300e-01 -5.09556174e-01 3.97131825e-03 -1.51005888e+00
4.05821770e-01 2.93931216e-02 -4.79727507e-01 4.06040877e-01
-1.24611191e-01 3.80684063e-02 4.87176448e-01 2.56602377e-01
-6.74545228e-01 1.05905676e+00 9.17492092e-01 -4.46843915e-02
-1.91358492e-01 1.44822106e-01 -2.82335013e-01 1.02557456e+00
9.16699946e-01 -1.29367739e-01 -6.64278805e-01 -7.08189785e-01
-3.13848943e-01 1.85500100e-01 3.16422433e-01 -1.14961493e+00
2.91260898e-01 6.90354034e-02 3.32854897e-01 -6.29629731e-01
7.27445602e-01 -9.03877556e-01 3.04070622e-01 6.57037079e-01
-1.52747467e-01 2.50234157e-01 2.06071690e-01 6.20801210e-01
-5.13454020e-01 -9.21310857e-03 8.76177430e-01 5.75027019e-02
-9.93612468e-01 5.75038314e-01 -6.54557347e-01 -4.67957735e-01
9.67385709e-01 -3.32340747e-01 -2.42563754e-01 -6.83923542e-01
-6.68579817e-01 2.89720327e-01 2.28481755e-01 5.62413752e-01
1.01871550e+00 -1.45482624e+00 -5.61511278e-01 -8.74426961e-03
-1.99695304e-01 -4.47550505e-01 6.55545533e-01 1.27395213e+00
-3.36302638e-01 5.94734907e-01 -2.50242114e-01 -5.00392973e-01
-1.52410018e+00 8.07926059e-01 4.11916435e-01 -4.27658588e-01
-6.77197695e-01 7.30487347e-01 7.51376569e-01 -4.19410430e-02
2.64695585e-01 -2.31343150e-01 -3.92556399e-01 -3.25513095e-01
6.28191292e-01 3.55558515e-01 -3.75794947e-01 -9.08567488e-01
-2.98432440e-01 5.99873900e-01 -9.69175547e-02 4.24401224e-01
1.41898751e+00 -3.10623467e-01 4.09586281e-02 3.82506967e-01
1.06348693e+00 -4.22366381e-01 -1.41960990e+00 -3.73993963e-01
-3.04458201e-01 -5.80512762e-01 2.42154166e-01 -1.77381963e-01
-1.37876868e+00 1.12930512e+00 6.16055787e-01 -2.09885798e-02
1.51253951e+00 -2.65943885e-01 9.52080727e-01 1.30685747e-01
1.93719551e-01 -8.28040600e-01 1.12881721e-03 7.14495301e-01
9.63036180e-01 -1.10993040e+00 -9.44378525e-02 -5.95836699e-01
-7.83042610e-01 1.03323710e+00 8.80303442e-01 -2.17678651e-01
7.41120458e-01 7.90102556e-02 -3.44534159e-01 2.07800791e-01
-1.12162220e+00 -5.97928911e-02 3.14329863e-01 4.42277551e-01
6.70887232e-01 -2.68487334e-01 -1.95911720e-01 7.15797365e-01
1.41940579e-01 3.54732066e-01 3.85396838e-01 4.66807187e-01
-3.85005474e-01 -7.85305560e-01 -1.13420390e-01 5.94790637e-01
-4.67204899e-01 -3.83647323e-01 1.84800848e-01 5.33702850e-01
-1.56705216e-01 8.97560060e-01 1.91679850e-01 -6.81310892e-01
1.24358229e-01 -4.89289351e-02 2.83860177e-01 -2.38489106e-01
-1.13591820e-01 4.56472039e-01 -6.79442985e-03 -7.65764236e-01
-6.60740495e-01 -4.08469468e-01 -9.08646822e-01 -3.04368198e-01
-3.45778972e-01 5.92736043e-02 1.82099849e-01 8.41259420e-01
4.81411666e-01 5.94496071e-01 5.22443593e-01 -1.05379927e+00
-3.98050845e-01 -9.52491522e-01 -3.77998024e-01 9.17009786e-02
2.10561559e-01 -7.85601974e-01 -8.01828802e-02 3.14159125e-01] | [8.900785446166992, 0.2868964672088623] |
68ac4983-1ce9-4437-84fc-dd20988170cb | improving-scientific-relation-classification | null | null | https://aclanthology.org/Y18-1015 | https://aclanthology.org/Y18-1015.pdf | Improving Scientific Relation Classification with Task Specific Supersense | null | ['Kentaro Inui', 'Paul Reisert', 'Naoya Inoue', 'Qin Dai'] | null | null | null | null | paclic-2018-12 | ['relation-classification'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.225683212280273, 3.765061855316162] |
c66d46df-8af5-431f-84a6-b8f0d1ea1b5c | mug-a-general-meeting-understanding-and | 2303.13939 | null | https://arxiv.org/abs/2303.13939v2 | https://arxiv.org/pdf/2303.13939v2.pdf | MUG: A General Meeting Understanding and Generation Benchmark | Listening to long video/audio recordings from video conferencing and online courses for acquiring information is extremely inefficient. Even after ASR systems transcribe recordings into long-form spoken language documents, reading ASR transcripts only partly speeds up seeking information. It has been observed that a range of NLP applications, such as keyphrase extraction, topic segmentation, and summarization, significantly improve users' efficiency in grasping important information. The meeting scenario is among the most valuable scenarios for deploying these spoken language processing (SLP) capabilities. However, the lack of large-scale public meeting datasets annotated for these SLP tasks severely hinders their advancement. To prompt SLP advancement, we establish a large-scale general Meeting Understanding and Generation Benchmark (MUG) to benchmark the performance of a wide range of SLP tasks, including topic segmentation, topic-level and session-level extractive summarization and topic title generation, keyphrase extraction, and action item detection. To facilitate the MUG benchmark, we construct and release a large-scale meeting dataset for comprehensive long-form SLP development, the AliMeeting4MUG Corpus, which consists of 654 recorded Mandarin meeting sessions with diverse topic coverage, with manual annotations for SLP tasks on manual transcripts of meeting recordings. To the best of our knowledge, the AliMeeting4MUG Corpus is so far the largest meeting corpus in scale and facilitates most SLP tasks. In this paper, we provide a detailed introduction of this corpus, SLP tasks and evaluation methods, baseline systems and their performance. | ['Zhou Zhao', 'Yi Ren', 'Jinglin Liu', 'Zhijie Yan', 'Wen Wang', 'Qian Chen', 'Hai Yu', 'Jiaqing Liu', 'Chong Deng', 'Qinglin Zhang'] | 2023-03-24 | null | null | null | null | ['topic-coverage', 'extractive-summarization', 'keyphrase-extraction'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 5.22670746e-01 4.14133817e-01 -1.30720139e-01 -2.57791370e-01
-1.92035615e+00 -8.94051373e-01 6.85508728e-01 4.09891725e-01
-2.65012264e-01 6.63025677e-01 7.35430419e-01 -2.08052173e-01
1.94000974e-02 -3.87757346e-02 -5.01304924e-01 -3.28363329e-01
-3.75273041e-02 5.61309576e-01 2.40111828e-01 -2.29293242e-01
6.60523534e-01 -6.94032162e-02 -1.53137529e+00 7.49484837e-01
1.07998514e+00 6.40895903e-01 5.28246105e-01 1.13624156e+00
-3.29069555e-01 5.68169177e-01 -1.33635485e+00 4.82530743e-02
-3.38155985e-01 -5.07795572e-01 -1.13568509e+00 3.17632705e-01
6.05039001e-01 -3.07521343e-01 -2.01648340e-01 4.39319968e-01
7.97337532e-01 2.10998341e-01 5.00587583e-01 -1.15811682e+00
-9.03724208e-02 1.26142597e+00 -4.56481844e-01 3.65562230e-01
1.08899915e+00 -1.56195328e-01 1.31231844e+00 -9.61588800e-01
6.63811564e-01 1.11873889e+00 2.60043621e-01 6.41242385e-01
-6.84807718e-01 -6.18207514e-01 1.68846518e-01 -1.14272028e-01
-1.47338450e+00 -9.55078781e-01 5.38350761e-01 -2.37828672e-01
1.28658438e+00 6.10589683e-01 5.80219448e-01 1.10291553e+00
-1.51988119e-01 1.70148146e+00 5.43256342e-01 -6.56550109e-01
1.49363931e-02 1.36083186e-01 5.16961575e-01 3.12298238e-01
-3.17855775e-01 -9.23173785e-01 -1.29235721e+00 -1.28535330e-01
3.05096865e-01 -6.61322176e-01 -7.39882171e-01 5.68407953e-01
-1.40602803e+00 4.17604834e-01 -5.87544382e-01 4.13928241e-01
-3.45059425e-01 -3.22549909e-01 4.36498553e-01 4.91497278e-01
6.09528303e-01 7.43988514e-01 -5.80204904e-01 -7.42154419e-01
-1.38212967e+00 5.56608856e-01 1.41179919e+00 1.25072527e+00
2.80376971e-01 -1.01770073e-01 -6.76033318e-01 1.15605652e+00
1.95979491e-01 4.42529947e-01 6.37212634e-01 -1.02345705e+00
1.01804578e+00 4.73361015e-01 -1.24323154e-02 -6.32787585e-01
-2.74471998e-01 5.21625839e-02 -4.54869390e-01 -8.14094186e-01
3.39441955e-01 -3.84984583e-01 -7.18889296e-01 1.41600215e+00
5.58310002e-02 7.65688047e-02 3.28202248e-01 4.20100927e-01
1.65645742e+00 1.41104388e+00 -2.49098539e-01 -7.19012499e-01
1.54533947e+00 -9.38452482e-01 -1.07834995e+00 -3.33377957e-01
6.24416888e-01 -1.05995536e+00 1.21487939e+00 6.90730870e-01
-1.55355811e+00 -1.27271533e-01 -8.11984360e-01 -1.86229363e-01
1.51534518e-02 2.40509182e-01 2.87496954e-01 3.76013130e-01
-1.24252880e+00 5.56674935e-02 -6.88542604e-01 -6.00625038e-01
1.12864524e-01 4.91573885e-02 -1.52329564e-01 -3.41958813e-02
-1.23602080e+00 3.00319791e-01 2.57729620e-01 -9.73808467e-02
-9.56851006e-01 -9.74897504e-01 -9.37595844e-01 2.07986191e-01
6.90624118e-01 -4.85960534e-03 2.09139228e+00 -3.43043059e-01
-1.71858895e+00 8.83449614e-01 -3.40901554e-01 -3.97995979e-01
9.26263928e-02 -5.32636702e-01 -3.41055691e-01 4.19285685e-01
9.90060046e-02 7.31112123e-01 6.17170453e-01 -7.70231068e-01
-8.13583910e-01 -1.15409411e-01 -2.68420547e-01 7.68804729e-01
-4.65893120e-01 5.07914126e-01 -6.76134884e-01 -5.23831725e-01
1.85494870e-01 -6.21870995e-01 2.21973151e-01 -1.09131110e+00
-7.97196984e-01 -7.96420872e-01 7.89694965e-01 -1.05115640e+00
1.83811700e+00 -2.23251414e+00 3.52813937e-02 -7.77440295e-02
6.47068173e-02 1.10865168e-01 -2.09228918e-01 9.41297591e-01
3.06052864e-01 3.25223595e-01 -7.47694867e-03 -5.06406486e-01
8.74420181e-02 -1.56375006e-01 -5.99998713e-01 -9.79611650e-02
4.99681197e-02 7.89434910e-01 -7.56710231e-01 -7.29164243e-01
-9.61244628e-02 1.12844199e-01 -4.00946110e-01 3.54104519e-01
-3.36508065e-01 3.42356354e-01 -5.96836269e-01 6.34843230e-01
3.85635011e-02 2.77376175e-03 -1.20487586e-01 3.41633379e-01
-3.79852414e-01 1.07550514e+00 -1.14022851e+00 1.98979926e+00
-3.15521866e-01 1.10376883e+00 3.88854563e-01 -7.47520685e-01
6.96500957e-01 9.45812285e-01 5.32169521e-01 -2.26508215e-01
-1.91872358e-01 1.65188521e-01 -3.41894299e-01 -4.97018278e-01
1.16895318e+00 4.82756227e-01 -6.84369326e-01 6.13225281e-01
3.62247169e-01 -8.20620179e-01 6.81378007e-01 9.48936522e-01
1.32019961e+00 -5.01861274e-01 5.02663374e-01 -1.78546757e-01
5.08673370e-01 1.12582505e-01 2.73244619e-01 9.09764886e-01
-7.27914870e-02 8.69313419e-01 5.96874535e-01 3.29753518e-01
-4.17989135e-01 -7.82624125e-01 2.76903138e-02 1.60415292e+00
-3.65428209e-01 -8.91428769e-01 -1.05840564e+00 -3.27904791e-01
-4.80731010e-01 9.21085000e-01 1.52652755e-01 2.87082791e-01
-7.39196658e-01 -4.10528094e-01 8.87092829e-01 1.85586572e-01
3.90829355e-01 -1.42307603e+00 -7.40205646e-02 3.45157027e-01
-8.36858451e-01 -1.46671462e+00 -9.51075077e-01 -1.67259976e-01
-4.99088049e-01 -8.21277678e-01 -7.72518218e-01 -1.02277982e+00
2.62601554e-01 4.93491679e-01 1.36927485e+00 -2.81887919e-01
-1.23025745e-01 7.73560345e-01 -5.41774809e-01 -7.79798388e-01
-7.34204412e-01 5.40431738e-01 -7.61389062e-02 -4.10095692e-01
4.08199638e-01 -2.51772612e-01 -1.93175852e-01 1.40822008e-01
-6.86186254e-01 1.79511845e-01 5.69798827e-01 2.83694029e-01
3.07378918e-01 -9.06308666e-02 8.60177159e-01 -8.21738601e-01
1.14065242e+00 -1.13371678e-01 -2.27507994e-01 3.42644244e-01
8.40205476e-02 -4.66427833e-01 3.85126509e-02 -3.49795014e-01
-1.32539439e+00 -3.38292003e-01 -3.63695949e-01 3.33821803e-01
-3.02510500e-01 8.25530231e-01 -3.33121121e-01 6.47368252e-01
5.69116056e-01 6.46827221e-01 -4.44012880e-01 -4.87115890e-01
3.76173966e-02 1.24424744e+00 5.99380553e-01 -5.58677912e-01
3.79352361e-01 -2.64212281e-01 -9.06087160e-01 -1.82539618e+00
-1.13785160e+00 -1.11845756e+00 -3.34644318e-01 -2.24029094e-01
7.59693027e-01 -1.19658947e+00 -5.72831988e-01 5.48831761e-01
-1.26870573e+00 -4.72548723e-01 -2.69781798e-01 3.45325291e-01
-5.57519436e-01 3.18421453e-01 -8.85106027e-01 -1.03033984e+00
-5.86824894e-01 -8.27019870e-01 1.33608592e+00 3.34391862e-01
-8.44151974e-01 -5.43054640e-01 2.61319466e-02 8.05718958e-01
-3.75861898e-02 -2.69652426e-01 3.89906108e-01 -1.17281985e+00
-4.49578315e-01 -2.75355577e-01 3.34665149e-01 1.66731626e-01
-1.22318372e-01 1.63527340e-01 -9.84195828e-01 -2.48364076e-01
-2.56012499e-01 -6.32753551e-01 8.24525833e-01 6.52855694e-01
9.90876675e-01 -5.66668868e-01 -2.08161116e-01 -4.01269421e-02
4.50624704e-01 1.59457788e-01 4.00907010e-01 -7.88323879e-02
5.00144839e-01 7.74167836e-01 7.66631782e-01 5.18229485e-01
6.24039590e-01 4.04233485e-01 -3.57338995e-01 4.24451351e-01
5.78484125e-02 -3.64946216e-01 1.00242317e+00 1.65019512e+00
1.09809183e-01 -7.30262637e-01 -1.10123706e+00 7.88824797e-01
-1.50922096e+00 -7.89429128e-01 3.85826826e-02 1.84202266e+00
1.40667796e+00 1.00176461e-01 4.83338870e-02 3.27229559e-01
4.78778958e-01 4.42798525e-01 2.05057133e-02 -3.14804465e-01
5.48562929e-02 1.94469665e-03 -1.14772119e-01 4.17677104e-01
-1.05710781e+00 1.13548219e+00 5.66597986e+00 1.06689703e+00
-6.09244108e-01 2.29706969e-02 4.08657968e-01 -2.61797339e-01
-1.42689243e-01 -4.76592720e-01 -1.44429350e+00 1.13904878e-01
1.33087647e+00 -6.08143210e-01 6.39827847e-02 6.98790431e-01
5.39463937e-01 -2.33458236e-01 -1.37151742e+00 8.97381246e-01
3.11668336e-01 -1.34313202e+00 3.16181295e-02 -1.57971039e-01
7.87786543e-01 -2.34902161e-03 -2.03650251e-01 7.14237630e-01
3.30342293e-01 -7.86366165e-01 5.87308228e-01 -1.84689686e-01
6.94397807e-01 -6.35754645e-01 5.13423502e-01 6.72992051e-01
-1.23241961e+00 2.75569856e-01 7.49877319e-02 4.19407263e-02
4.67267036e-01 3.82230788e-01 -1.27717710e+00 3.24853927e-01
6.66420221e-01 5.42506039e-01 -1.87119752e-01 8.27847362e-01
-3.74705404e-01 1.21801758e+00 -4.42232698e-01 -9.20799747e-02
3.91361952e-01 2.29964703e-01 1.01309311e+00 1.61612594e+00
3.49179417e-01 5.30797660e-01 5.16713023e-01 2.74769902e-01
-3.66828501e-01 4.44371730e-01 -4.53134090e-01 -5.90110302e-01
8.72546852e-01 1.31830812e+00 -7.59772420e-01 -5.17008007e-01
-3.85310501e-01 7.33249962e-01 -2.35086113e-01 5.60371459e-01
-3.31844330e-01 -4.63427693e-01 5.01713276e-01 8.65118802e-02
-1.50634497e-01 -3.64856154e-01 9.18224677e-02 -1.16411436e+00
-1.22077474e-02 -1.27757847e+00 5.12348175e-01 -4.71464574e-01
-7.80956388e-01 3.99232537e-01 2.30871752e-01 -9.05407190e-01
-6.57329857e-01 5.95410131e-02 -8.11621606e-01 5.17603397e-01
-1.20565796e+00 -7.55587220e-01 -2.04835415e-01 3.69661152e-01
1.76572585e+00 -1.61903188e-01 7.20192790e-01 8.16598013e-02
-7.81794548e-01 6.03243232e-01 -3.21515083e-01 1.32544518e-01
9.46259022e-01 -1.30574560e+00 3.58447671e-01 7.20844686e-01
3.90270263e-01 5.42240202e-01 9.09098268e-01 -5.30159891e-01
-1.34736490e+00 -8.02691162e-01 1.25382185e+00 -5.92704833e-01
6.90311015e-01 -5.89690685e-01 -1.10316277e+00 7.50867188e-01
5.90496838e-01 -9.39003110e-01 8.06368768e-01 1.34460598e-01
3.11737329e-01 1.84383661e-01 -5.66499174e-01 5.52600682e-01
8.16787601e-01 -4.78089392e-01 -1.03825164e+00 5.78970253e-01
1.15519667e+00 -7.73320138e-01 -8.08635771e-01 2.09900409e-01
1.85587838e-01 -3.01739573e-01 7.06944048e-01 -2.23379254e-01
5.76578259e-01 1.16036467e-01 -9.22394218e-04 -1.34787989e+00
3.99502635e-01 -1.48984134e+00 -1.46321386e-01 1.96169007e+00
7.74854004e-01 -1.86129794e-01 6.39873922e-01 6.19337380e-01
-6.99786365e-01 -3.39840055e-01 -7.72432625e-01 -2.94342697e-01
-2.67318666e-01 -6.03909612e-01 1.83122382e-01 8.36365104e-01
6.05749786e-01 8.38915288e-01 -1.32058427e-01 8.14691931e-02
3.36808503e-01 -6.55523837e-02 9.29509938e-01 -1.15364707e+00
-1.88958389e-03 -3.91170591e-01 2.81128764e-01 -1.46185434e+00
2.69041151e-01 -6.29094303e-01 6.12615228e-01 -1.77094936e+00
2.02628762e-01 7.80973136e-02 1.07925579e-01 4.46816951e-01
-1.31521821e-01 -2.10337043e-01 3.63986157e-02 1.63912430e-01
-1.03869760e+00 3.12884837e-01 1.06726086e+00 -1.02240510e-01
-8.53344202e-01 3.76205623e-01 -7.10489869e-01 9.05110896e-01
5.97027659e-01 -1.28873110e-01 -4.92717862e-01 -3.00121039e-01
1.22192167e-01 5.92062533e-01 -3.79771709e-01 -7.41158545e-01
6.30320251e-01 -8.02491978e-02 -1.55822605e-01 -1.20135927e+00
2.63317257e-01 -7.36168325e-02 -6.76577389e-01 -2.39464313e-01
-7.90114224e-01 -2.22059488e-01 4.31033254e-01 1.46384463e-01
-7.43708909e-01 -1.81846261e-01 1.14913270e-01 -2.08835378e-01
-8.50440025e-01 9.01394114e-02 -1.09477687e+00 6.24662101e-01
6.31919384e-01 -2.15616569e-01 -2.33519554e-01 -8.97814512e-01
-4.86120731e-01 7.85036325e-01 -1.32110864e-01 4.78914469e-01
7.48517871e-01 -7.20619678e-01 -1.08811331e+00 7.42755011e-02
1.24231152e-01 5.23170412e-01 3.44781607e-01 8.10509741e-01
-2.27175295e-01 7.90015817e-01 3.70965123e-01 -6.61622107e-01
-1.75111258e+00 -2.96460062e-01 -4.26923364e-01 -5.12272775e-01
-6.75958991e-01 1.11289871e+00 3.80371898e-01 -3.08261037e-01
6.53313398e-01 -5.58514714e-01 -6.17950618e-01 5.45890093e-01
1.04634535e+00 3.35518599e-01 1.76032826e-01 -3.60845953e-01
-1.14783466e-01 2.50016209e-02 -3.33121717e-01 -5.52460849e-01
1.30346513e+00 -5.24284720e-01 -5.16212247e-02 7.19247758e-01
7.60465324e-01 3.05956155e-01 -9.23372924e-01 -2.45540872e-01
3.99942219e-01 1.02801047e-01 7.92109296e-02 -6.87667727e-01
-3.42025101e-01 7.10800529e-01 -4.55051631e-01 5.18060803e-01
1.07758677e+00 3.78564686e-01 1.00296605e+00 7.76037574e-01
1.49069205e-01 -1.34754658e+00 4.32709783e-01 9.49057341e-01
1.28826940e+00 -1.26388907e+00 -8.99991617e-02 -6.80971026e-01
-8.50983918e-01 7.70609915e-01 6.71981096e-01 4.66085225e-01
2.30065137e-01 3.84592742e-01 6.54143021e-02 -2.02551857e-01
-1.00956237e+00 1.88153803e-01 4.85320330e-01 1.67345315e-01
7.43251204e-01 6.16139024e-02 -1.54214695e-01 8.45410764e-01
-7.83328593e-01 -2.93264598e-01 9.91346478e-01 9.21840966e-01
-7.82199085e-01 -7.59673059e-01 -3.93344045e-01 3.56215000e-01
-8.93247306e-01 -3.37645859e-01 -7.33289957e-01 7.03756809e-01
-9.14627492e-01 1.28597641e+00 9.50461924e-02 4.27766033e-02
3.78980905e-01 4.22963887e-01 1.12740636e-01 -1.37004232e+00
-6.30207419e-01 4.19501573e-01 5.97977638e-01 -4.79271829e-01
-3.23937237e-01 -7.95139790e-01 -1.30234170e+00 -9.64563191e-02
-3.20079207e-01 6.44434631e-01 5.69481969e-01 9.28953171e-01
2.32440859e-01 6.72150433e-01 3.35394651e-01 -9.29583609e-01
-4.66493040e-01 -1.55600679e+00 -5.63166142e-01 -4.90215383e-02
1.62928089e-01 1.23733915e-01 -2.79377609e-01 3.05736303e-01] | [12.610821723937988, 9.392475128173828] |
2426c2c3-5dde-4ad9-ae9f-0126c324911f | rethinking-the-evaluation-for-conversational | 2305.13112 | null | https://arxiv.org/abs/2305.13112v1 | https://arxiv.org/pdf/2305.13112v1.pdf | Rethinking the Evaluation for Conversational Recommendation in the Era of Large Language Models | The recent success of large language models (LLMs) has shown great potential to develop more powerful conversational recommender systems (CRSs), which rely on natural language conversations to satisfy user needs. In this paper, we embark on an investigation into the utilization of ChatGPT for conversational recommendation, revealing the inadequacy of the existing evaluation protocol. It might over-emphasize the matching with the ground-truth items or utterances generated by human annotators, while neglecting the interactive nature of being a capable CRS. To overcome the limitation, we further propose an interactive Evaluation approach based on LLMs named iEvaLM that harnesses LLM-based user simulators. Our evaluation approach can simulate various interaction scenarios between users and systems. Through the experiments on two publicly available CRS datasets, we demonstrate notable improvements compared to the prevailing evaluation protocol. Furthermore, we emphasize the evaluation of explainability, and ChatGPT showcases persuasive explanation generation for its recommendations. Our study contributes to a deeper comprehension of the untapped potential of LLMs for CRSs and provides a more flexible and easy-to-use evaluation framework for future research endeavors. The codes and data are publicly available at https://github.com/RUCAIBox/iEvaLM-CRS. | ['Ji-Rong Wen', 'Jingyuan Wang', 'Wayne Xin Zhao', 'Xinyu Tang', 'Xiaolei Wang'] | 2023-05-22 | null | null | null | null | ['explanation-generation'] | ['natural-language-processing'] | [-1.56954855e-01 5.27859151e-01 -3.68444137e-02 -3.90172452e-01
-5.83937705e-01 -5.89192629e-01 8.23393583e-01 -1.74772292e-01
9.34929848e-02 6.98337436e-01 6.08730316e-01 -7.55868256e-01
-2.31091216e-01 -5.87177336e-01 -3.80722076e-01 -1.57509923e-01
-3.15125845e-02 4.78769332e-01 -1.07109509e-01 -6.56842232e-01
3.68343949e-01 -3.95544991e-02 -1.67621017e+00 7.78702617e-01
9.56312120e-01 3.74208421e-01 5.01244485e-01 9.11867201e-01
-2.07703397e-01 9.19163525e-01 -5.38184583e-01 -6.98093355e-01
-1.23672180e-01 -4.33685780e-01 -1.11794674e+00 -1.41979873e-01
5.06506162e-03 -3.52596164e-01 -2.53324248e-02 4.28693622e-01
4.90835756e-01 4.16787148e-01 2.95953482e-01 -1.45469594e+00
-1.08998418e+00 1.15089071e+00 2.06277311e-01 -1.30170032e-01
9.52212393e-01 1.41665088e-02 1.24729002e+00 -8.43590856e-01
5.25590837e-01 1.20252883e+00 5.84508717e-01 7.64732659e-01
-8.74344885e-01 -5.06319165e-01 2.82913804e-01 1.56441987e-01
-1.19621897e+00 -3.11377376e-01 4.00162041e-01 -3.58024895e-01
9.37709153e-01 9.84613717e-01 4.98500526e-01 1.46282208e+00
-1.26548097e-01 7.90795147e-01 1.04251850e+00 -5.09546340e-01
1.37591392e-01 8.22724640e-01 4.75229621e-01 5.62903464e-01
-7.05323443e-02 6.97164536e-02 -4.77560222e-01 -3.64002615e-01
5.16228080e-01 -4.57864255e-02 -4.11304057e-01 7.22522289e-02
-9.46567595e-01 8.29780400e-01 4.29203101e-02 5.49599648e-01
-2.95386016e-01 -2.66276360e-01 1.21727072e-01 4.81885105e-01
5.43250501e-01 7.14231491e-01 -3.31198096e-01 -6.52924061e-01
-4.29940671e-01 3.82594913e-01 1.39156878e+00 1.06511915e+00
2.79469252e-01 -3.64593327e-01 -2.89049238e-01 9.24203873e-01
6.45441234e-01 3.48197341e-01 4.27132815e-01 -1.04244006e+00
1.10912681e-01 6.38567269e-01 4.49959576e-01 -1.10400629e+00
-3.39103460e-01 -3.94933909e-01 -3.26870888e-01 -2.76140302e-01
2.57676125e-01 -2.21491635e-01 5.42696565e-02 1.52757370e+00
1.16736405e-01 2.37615556e-01 2.46947065e-01 1.01749325e+00
1.22235096e+00 6.19478762e-01 1.52521074e-01 -2.66050249e-02
1.37610221e+00 -1.25034845e+00 -6.76310003e-01 2.48758465e-01
7.88643837e-01 -8.34872246e-01 1.76470077e+00 4.22121584e-01
-9.58204210e-01 -4.65489417e-01 -7.46232390e-01 1.48050562e-01
-2.86620975e-01 1.15991659e-01 9.60767865e-01 8.41650844e-01
-1.09232879e+00 3.31959993e-01 -5.05866766e-01 -7.34807789e-01
-3.38790417e-01 1.48088157e-01 7.47846290e-02 1.82487488e-01
-1.43468225e+00 8.83239627e-01 -2.36662582e-01 1.26640603e-01
-6.23546660e-01 -5.06240726e-01 -5.30717850e-01 -1.50920707e-03
4.56314266e-01 -5.78317583e-01 1.81899691e+00 -7.97864616e-01
-1.91099465e+00 3.12394023e-01 -5.35291024e-02 -3.66335481e-01
3.60045850e-01 -3.03199142e-01 -6.81777000e-01 -3.21290076e-01
-1.25104412e-01 2.17697933e-01 -2.43110396e-02 -1.45878816e+00
-4.02968466e-01 2.88439304e-01 7.74185777e-01 3.06051821e-01
-2.94154018e-01 1.41217157e-01 -4.60689425e-01 -2.43627802e-01
-5.56575656e-01 -1.08530986e+00 -3.85735214e-01 -6.04930162e-01
-4.07094479e-01 -3.24063331e-01 3.59375656e-01 -3.62391710e-01
1.63319755e+00 -1.74078727e+00 -1.53164461e-01 1.18957803e-01
1.94949403e-01 4.98143226e-01 -3.66255999e-01 1.26455593e+00
3.51811975e-01 4.25236225e-01 3.27348024e-01 -6.83048785e-01
3.06970984e-01 1.84530839e-01 -3.19017500e-01 -1.24126405e-01
-4.16162461e-01 7.76988208e-01 -9.42751288e-01 -1.84823871e-01
3.56315136e-01 6.61757588e-01 -6.91288710e-01 6.19516850e-01
-3.03755313e-01 5.26957214e-01 -5.81919491e-01 1.86052188e-01
3.14144373e-01 -6.05662405e-01 3.09676081e-01 6.86274469e-02
-3.92678946e-01 7.61145711e-01 -1.08346105e+00 1.52738059e+00
-8.10879171e-01 4.37489837e-01 -1.61033750e-01 -2.67178863e-01
8.06817055e-01 6.41149521e-01 1.67009205e-01 -5.44580340e-01
2.25950465e-01 2.03684513e-02 3.29489224e-02 -7.23419726e-01
8.48753512e-01 3.07258248e-01 -1.94949172e-02 9.70366716e-01
-2.03312829e-01 2.01130271e-01 7.67661780e-02 6.10714614e-01
8.58945787e-01 1.16441585e-01 3.30776900e-01 -2.31502205e-01
7.16289997e-01 -1.53902382e-01 -6.60199225e-02 1.10272384e+00
2.90314686e-02 3.79780531e-01 1.92747906e-01 -2.28731066e-01
-5.23669720e-01 -5.72158813e-01 3.10471971e-02 1.35515964e+00
2.12959141e-01 -9.63851213e-01 -8.95999014e-01 -6.59193635e-01
-3.10182065e-01 1.24450171e+00 -4.60914940e-01 2.66662121e-01
-7.17754215e-02 -4.86721158e-01 5.34713149e-01 1.04598723e-01
1.92221895e-01 -1.25010550e+00 -3.20621431e-01 1.25307977e-01
-6.42689586e-01 -1.13848686e+00 -3.26554000e-01 -5.06478548e-01
-4.57477182e-01 -9.42739487e-01 -2.76130497e-01 -4.22800928e-01
3.63905132e-01 7.17369437e-01 1.33622348e+00 7.25419521e-01
3.15508634e-01 8.44609916e-01 -9.60727751e-01 -2.79961467e-01
-7.07401693e-01 1.86423704e-01 7.51499906e-02 -1.81084320e-01
4.85032439e-01 -7.25213766e-01 -7.33365834e-01 8.52672696e-01
-4.81656879e-01 6.07079208e-01 3.00009102e-01 3.33950460e-01
-8.07008594e-02 -5.19592524e-01 8.16757202e-01 -1.41372335e+00
1.26681542e+00 -8.39328110e-01 -1.01710647e-01 2.71479428e-01
-9.40808058e-01 -1.69510812e-01 5.20656645e-01 -3.79279822e-01
-1.29465616e+00 -7.33117223e-01 -5.47748506e-01 3.43570173e-01
-3.17839116e-01 7.30107486e-01 1.82045281e-01 9.40040052e-02
7.30585098e-01 -5.71734942e-02 -1.70978069e-01 -5.79198539e-01
6.78210616e-01 1.12406611e+00 1.25614762e-01 -8.02741051e-01
4.69623148e-01 5.84369190e-02 -8.92565191e-01 -7.38825500e-01
-7.14373529e-01 -4.89670962e-01 2.92092692e-02 -6.07451737e-01
4.61578339e-01 -7.66445756e-01 -1.19045877e+00 -2.85043865e-01
-1.12796903e+00 -5.40867150e-01 7.11416155e-02 4.82494056e-01
-3.18075836e-01 5.80601394e-01 -8.07082534e-01 -1.29667723e+00
-6.61072493e-01 -9.48683798e-01 7.49348283e-01 2.10837498e-01
-8.45538020e-01 -1.13315785e+00 2.73393959e-01 9.62702394e-01
7.70463943e-01 -3.26900482e-01 5.47859430e-01 -1.12148952e+00
-4.37517881e-01 -1.92797318e-01 -2.89934129e-02 4.08435613e-02
-1.34348288e-01 1.35690331e-01 -1.02415633e+00 -1.52767971e-01
-2.36057371e-01 -2.63831317e-01 -7.17743710e-02 -6.97626993e-02
9.59576547e-01 -5.23805797e-01 -2.12039337e-01 -4.94934879e-02
1.00595713e+00 1.39529884e-01 7.15134203e-01 3.36017668e-01
4.26746309e-01 7.13140547e-01 6.74339533e-01 5.81728458e-01
9.08442378e-01 7.63293087e-01 2.28767887e-01 -3.16795520e-02
-8.05537701e-02 -5.45414865e-01 4.73360926e-01 1.29273081e+00
-5.20342290e-01 -8.01897347e-01 -6.04644179e-01 2.76814103e-01
-2.08211756e+00 -8.89150858e-01 -4.20923859e-01 2.13733482e+00
5.11259258e-01 -1.08337812e-01 7.24559592e-04 -2.24050477e-01
3.62568319e-01 -9.07933339e-02 8.15486833e-02 -7.42856443e-01
2.46168807e-01 -1.47365838e-01 -2.47919291e-01 9.26095605e-01
-3.46546739e-01 7.46467292e-01 6.17559242e+00 5.55727124e-01
-6.69124722e-01 3.02901626e-01 3.28113139e-01 1.88950703e-01
-6.73168957e-01 7.17655495e-02 -7.58102059e-01 4.19934750e-01
1.50939846e+00 -2.72180170e-01 6.34672880e-01 8.44196975e-01
7.64548779e-01 1.64596483e-01 -1.11452198e+00 5.63044965e-01
3.69181558e-02 -1.35853422e+00 2.87473816e-02 1.33473441e-01
5.42293251e-01 3.81895341e-02 -1.07349671e-01 7.86217213e-01
5.62677205e-01 -8.25781226e-01 4.66877043e-01 4.53031868e-01
2.71027595e-01 -3.54130298e-01 1.02791381e+00 6.40813410e-01
-9.75889385e-01 1.17857888e-01 -1.79828346e-01 -5.87284863e-01
3.33742172e-01 2.68352032e-02 -1.20381296e+00 6.04672015e-01
4.36511189e-01 4.22064573e-01 -2.96825379e-01 6.24077260e-01
-2.58762717e-01 8.65004778e-01 4.06847596e-02 -4.64659750e-01
6.35655373e-02 -3.46439540e-01 5.22097647e-01 1.53007317e+00
3.53602409e-01 3.92886460e-01 2.80529797e-01 8.35269511e-01
5.07625341e-02 5.71504056e-01 -4.73172009e-01 -1.06625862e-01
7.29593635e-01 1.47472334e+00 -3.01474899e-01 -2.00274989e-01
-6.54573143e-01 7.50970125e-01 2.92880923e-01 3.54969203e-01
-8.82000625e-01 2.74119347e-01 6.20843232e-01 1.71409711e-01
-1.79367885e-01 2.47835703e-02 -2.52846867e-01 -1.20409095e+00
-3.24234456e-01 -1.27665997e+00 1.42442405e-01 -9.52241719e-01
-1.23972464e+00 1.03397298e+00 6.65734112e-02 -1.13297844e+00
-3.84888798e-01 -1.85527802e-01 -8.75126898e-01 7.48494565e-01
-1.12823534e+00 -1.25017166e+00 -4.15582925e-01 4.07326013e-01
8.12524378e-01 -8.18045661e-02 1.24787343e+00 4.92433161e-01
-5.79022110e-01 6.82227135e-01 -1.70277163e-01 -4.09726471e-01
5.97750425e-01 -1.13863826e+00 4.53498781e-01 4.49439049e-01
2.65813649e-01 1.09187961e+00 1.22565210e+00 -4.52766836e-01
-1.31049037e+00 -6.78015113e-01 1.18437886e+00 -8.45840335e-01
6.63235962e-01 -5.18888414e-01 -9.39695299e-01 6.86296523e-01
6.54370606e-01 -8.35075498e-01 1.21086729e+00 7.86357582e-01
3.16596031e-02 4.38285708e-01 -9.14430499e-01 9.42087710e-01
1.04450703e+00 -5.24426758e-01 -5.90063453e-01 3.87474716e-01
8.80472064e-01 -2.22418442e-01 -9.27863121e-01 7.71157537e-03
8.76855016e-01 -1.09173822e+00 6.72366321e-01 -7.26187468e-01
3.99577022e-01 -1.20935678e-01 -1.97346181e-01 -1.20167613e+00
-1.53913379e-01 -9.46948111e-01 -9.65219215e-02 1.55600798e+00
8.83777440e-01 -5.64325631e-01 5.61994135e-01 1.16300905e+00
-3.22389692e-01 -7.36757696e-01 -1.79844201e-01 -3.41453761e-01
-3.36146295e-01 -8.01437080e-01 7.12858498e-01 9.17430937e-01
6.49024427e-01 4.65059906e-01 -6.99688196e-01 2.18957543e-01
1.17064223e-01 1.67172536e-01 1.17782342e+00 -1.14323306e+00
-6.40876174e-01 -2.06485078e-01 3.37615132e-01 -1.17787659e+00
-5.30622574e-03 -8.36535394e-01 1.74280035e-03 -1.56841242e+00
3.38019669e-01 -5.11207998e-01 -1.47869438e-01 3.13166469e-01
1.16671948e-02 1.83611259e-01 1.56709671e-01 1.76944673e-01
-1.07568753e+00 4.09299105e-01 1.21826530e+00 4.56696093e-01
-4.70025748e-01 3.35454315e-01 -1.26020813e+00 7.25898206e-01
8.75166059e-01 -3.60506743e-01 -7.58250058e-01 -1.29554719e-01
4.08325016e-01 2.51121521e-01 1.05257131e-01 -6.02641523e-01
1.57848179e-01 -1.38092488e-01 -5.13293684e-01 -1.51440084e-01
2.92477161e-01 -7.10636377e-01 4.71264929e-01 -3.75346057e-02
-7.96089411e-01 3.58329192e-02 1.44204676e-01 3.76750499e-01
1.15543723e-01 -2.52913862e-01 -1.06866345e-01 -4.24380749e-02
-4.95801717e-01 -1.34345457e-01 -8.71459723e-01 -4.04941648e-01
6.28193080e-01 -1.56706832e-02 -5.28892636e-01 -1.03728604e+00
-6.19795799e-01 3.59370291e-01 2.50740618e-01 7.76522815e-01
5.00234067e-01 -1.08354950e+00 -7.44296908e-01 -8.87473747e-02
2.87333041e-01 -8.49752724e-01 4.24250782e-01 7.56117225e-01
-1.63591608e-01 5.90246499e-01 1.19216226e-01 -1.97329804e-01
-1.51302207e+00 3.35493803e-01 1.37524232e-01 -2.86766708e-01
-5.53912878e-01 5.68231106e-01 3.43801528e-02 -8.43219161e-01
3.19047570e-01 -2.67472297e-01 -5.83993495e-01 -4.53025997e-01
7.67192781e-01 3.90888810e-01 -1.18360683e-01 -3.69825393e-01
-9.84440297e-02 -1.79057494e-01 -1.01709239e-01 -1.99455619e-01
1.19495118e+00 -6.93829119e-01 3.09199039e-02 6.15535736e-01
6.89983428e-01 4.95027512e-01 -7.17057288e-01 -2.21485630e-01
-5.59776537e-02 -4.34935480e-01 -2.07508847e-01 -1.27937365e+00
-5.19667685e-01 4.45895106e-01 3.69957060e-01 7.87750244e-01
5.85776150e-01 4.44744620e-03 6.39181793e-01 4.85672325e-01
5.81879735e-01 -6.73327923e-01 -1.52792409e-01 3.04201931e-01
1.15113270e+00 -1.28484488e+00 -2.56035537e-01 -5.03257632e-01
-1.12660599e+00 9.37867939e-01 6.55421138e-01 2.32557386e-01
5.82067549e-01 5.61350733e-02 3.28248113e-01 -3.34075749e-01
-1.28976107e+00 -8.95957649e-02 3.66231948e-01 3.15917671e-01
1.24568713e+00 3.43177348e-01 -8.24548066e-01 1.29330707e+00
-4.81180876e-01 1.77613676e-01 6.81373060e-01 4.65395898e-01
-3.48714024e-01 -1.38348198e+00 -6.11706376e-02 1.46281287e-01
-3.38597119e-01 -2.62395322e-01 -5.96248209e-01 8.16532671e-01
-1.85008481e-01 1.75745618e+00 -5.21044552e-01 -7.91174769e-01
5.51124454e-01 -3.31850126e-02 -1.16203442e-01 -6.79901600e-01
-1.03576303e+00 -2.38641337e-01 8.08664441e-01 -5.80602586e-01
-5.21512687e-01 -4.71109539e-01 -1.05631173e+00 -7.29111075e-01
-5.08846998e-01 8.72015357e-01 5.44257641e-01 9.47954416e-01
5.46882331e-01 3.51323336e-01 4.99875069e-01 -6.68321073e-01
-3.12130868e-01 -1.20113850e+00 -2.01368496e-01 3.95764649e-01
-1.29734784e-01 -4.53667283e-01 -3.94621521e-01 -1.87177047e-01] | [12.267576217651367, 7.533604145050049] |
3cff48d3-641f-43db-b30a-048ee9829ed0 | obstacle-transformer-a-trajectory-prediction | 2304.07711 | null | https://arxiv.org/abs/2304.07711v1 | https://arxiv.org/pdf/2304.07711v1.pdf | Obstacle-Transformer: A Trajectory Prediction Network Based on Surrounding Trajectories | Recurrent Neural Network, Long Short-Term Memory, and Transformer have made great progress in predicting the trajectories of moving objects. Although the trajectory element with the surrounding scene features has been merged to improve performance, there still exist some problems to be solved. One is that the time series processing models will increase the inference time with the increase of the number of prediction sequences. Another lies in which the features can not be extracted from the scene's image and point cloud in some situations. Therefore, this paper proposes an Obstacle-Transformer to predict trajectory in a constant inference time. An ``obstacle'' is designed by the surrounding trajectory rather than images or point clouds, making Obstacle-Transformer more applicable in a wider range of scenarios. Experiments are conducted on ETH and UCY data sets to verify the performance of our model. | ['ChengWei Wu', 'Quanqi Zhang', 'Qingjie Chai', 'Wendong Zhang'] | 2023-04-16 | null | null | null | null | ['trajectory-prediction'] | ['computer-vision'] | [-8.00419748e-02 -6.78597331e-01 -2.11883914e-02 -4.25846905e-01
-6.05503470e-02 -1.16790220e-01 5.12398362e-01 -1.72186688e-01
-5.12768388e-01 4.72195894e-01 -1.15025928e-02 -1.77724034e-01
-7.59158507e-02 -8.81521881e-01 -6.18798852e-01 -6.92849100e-01
-2.29982242e-01 4.03301306e-02 7.24698067e-01 -1.55740663e-01
2.66211271e-01 4.06911939e-01 -1.73212636e+00 3.47560287e-01
8.56172502e-01 1.07201409e+00 7.90308714e-01 4.65483427e-01
-4.13024396e-01 9.96437609e-01 -3.56345385e-01 -1.51570648e-01
3.87626320e-01 -1.33177578e-01 -5.08290470e-01 -2.52207845e-01
-2.11506009e-01 -3.76482874e-01 -6.23989403e-01 9.22110558e-01
3.90564561e-01 4.48669612e-01 3.26004088e-01 -1.37164581e+00
-3.58344555e-01 4.65296805e-01 -4.12302375e-01 4.45155203e-01
8.31217989e-02 1.46891192e-01 3.59121978e-01 -5.85036457e-01
5.22514105e-01 1.10758042e+00 8.10635746e-01 2.30077922e-01
-4.16658163e-01 -7.00841188e-01 5.15137911e-01 8.10912132e-01
-1.56379545e+00 -1.78580076e-01 8.10912549e-01 -4.64099735e-01
1.14496589e+00 2.11486563e-01 7.61808574e-01 1.12676036e+00
3.23947877e-01 7.57347345e-01 5.35557866e-01 9.45850555e-03
5.77612780e-02 1.63185820e-01 9.10133310e-03 3.08849216e-01
-1.86550051e-01 3.64380389e-01 -6.62943199e-02 2.29873061e-01
7.30585158e-01 4.66289252e-01 -2.01684594e-01 6.28935099e-02
-1.28615010e+00 5.98492384e-01 5.42881191e-01 4.31713164e-01
-3.59250128e-01 4.67190295e-02 5.68998694e-01 7.11193085e-02
4.70491827e-01 3.52422930e-02 -3.81603360e-01 -4.93641376e-01
-8.46766412e-01 3.00386667e-01 4.65178341e-01 1.07410038e+00
4.94716674e-01 1.78144231e-01 7.69301280e-02 6.02698922e-01
1.33183599e-01 6.16510153e-01 7.17549801e-01 -6.37009799e-01
6.76504910e-01 6.13471270e-01 2.50502884e-01 -1.52266526e+00
-5.81408620e-01 -3.79154414e-01 -1.02800584e+00 -5.54209836e-02
2.37625629e-01 -1.78988233e-01 -8.73451352e-01 1.55008137e+00
2.11410567e-01 7.57393837e-01 4.51193452e-02 1.09184921e+00
6.76938117e-01 1.36223114e+00 3.36174630e-02 -2.83492774e-01
9.08236027e-01 -8.49913180e-01 -7.14054346e-01 -1.39540702e-01
6.17704988e-01 -7.21573532e-01 7.22375274e-01 1.82379082e-01
-6.43915415e-01 -1.02383590e+00 -8.80547285e-01 -6.37328764e-03
-6.26606405e-01 1.69441074e-01 5.79766095e-01 1.07419610e-01
-6.85931385e-01 8.05486619e-01 -1.02855575e+00 -3.08117449e-01
1.92473054e-01 2.79481888e-01 3.68309952e-02 2.25459814e-01
-1.41108060e+00 1.03151226e+00 5.37437737e-01 5.51250398e-01
-6.07535064e-01 -4.39664423e-01 -6.43114448e-01 -3.14019062e-02
2.45061722e-02 -4.84321535e-01 1.10210669e+00 -8.42637658e-01
-1.33917403e+00 2.47388855e-01 -3.99081975e-01 -5.20254076e-01
7.76295900e-01 -2.44842827e-01 -6.95672214e-01 -1.74710542e-01
-3.00042816e-02 5.08900225e-01 5.70095003e-01 -7.80879974e-01
-1.07883537e+00 -3.42380762e-01 -1.78573765e-02 2.99684852e-01
-2.64612645e-01 -5.03012910e-02 -5.15642166e-01 -4.11673814e-01
6.81938380e-02 -1.20413685e+00 -6.51784122e-01 -2.27735490e-01
-2.68862583e-02 -3.97629142e-01 1.30611289e+00 -5.36977291e-01
1.28103864e+00 -2.24460721e+00 -8.80003944e-02 -3.48172560e-02
-9.63631719e-02 4.26855326e-01 -1.51837140e-01 4.28999662e-01
-3.53390798e-02 3.96516174e-02 1.68313235e-01 -1.71655476e-01
-3.62834394e-01 1.73462704e-01 -7.73813903e-01 3.30035955e-01
-1.01511799e-01 7.65294790e-01 -8.02711904e-01 -5.95901668e-01
4.13359016e-01 4.92833078e-01 -1.82220712e-01 3.65205780e-02
-2.17696905e-01 5.54192007e-01 -6.81872904e-01 1.02423996e-01
6.99353695e-01 -1.58434972e-01 -4.45782483e-01 -6.92652864e-03
-6.09344900e-01 1.02387607e-01 -1.08279860e+00 1.69525850e+00
-2.32382342e-01 7.02537656e-01 -5.79196990e-01 -9.33403432e-01
8.17642868e-01 3.36935550e-01 6.76274598e-01 -7.88756132e-01
1.60316303e-01 -1.66485235e-02 -4.25536893e-02 -1.05803180e+00
6.25246108e-01 2.47383751e-02 1.05665542e-01 1.26404732e-01
-4.51559991e-01 2.75863051e-01 1.94484800e-01 -1.23469912e-01
7.53078401e-01 4.21653152e-01 -1.14717275e-01 1.13487050e-01
4.12291735e-01 2.76391059e-01 9.90572631e-01 4.72902298e-01
-1.98255688e-01 4.86319333e-01 -6.22239616e-03 -7.39787102e-01
-1.14522874e+00 -7.77859867e-01 -7.95689300e-02 7.89029002e-01
5.79766393e-01 -4.13889557e-01 -3.96004379e-01 -2.70984292e-01
-3.82943481e-01 7.17777312e-01 -2.71505862e-01 -1.78015292e-01
-7.84080625e-01 -6.78326666e-01 4.39491510e-01 6.93773687e-01
7.52516329e-01 -1.04611552e+00 -1.09802222e+00 4.06997949e-01
-5.11835575e-01 -1.29277217e+00 -2.15780407e-01 -3.36783290e-01
-9.52797055e-01 -7.61447906e-01 -4.02190745e-01 -6.83992267e-01
4.94293869e-01 5.00838757e-01 5.90830803e-01 -2.49792024e-01
-7.03209266e-02 -1.27350241e-01 -4.85866845e-01 -7.04385996e-01
8.35787207e-02 5.22924401e-03 1.10376634e-01 -2.11103082e-01
4.69776094e-01 -6.75697625e-01 -5.09176850e-01 4.39451128e-01
-6.77597702e-01 4.04939324e-01 3.65830928e-01 4.26640421e-01
4.93987858e-01 7.44677007e-01 4.01185930e-01 -4.84312445e-01
3.55514079e-01 -4.99367535e-01 -7.44033277e-01 8.49646032e-02
-3.63947511e-01 -2.41162255e-01 9.43739295e-01 -6.76262796e-01
-1.14083326e+00 1.45819392e-02 -2.51468331e-01 -8.17800641e-01
-1.75958872e-01 7.31013000e-01 2.76444666e-02 2.46875346e-01
3.41972768e-01 4.84825462e-01 -1.98042661e-01 -2.56599069e-01
7.11060613e-02 6.19672000e-01 3.46342415e-01 -1.12503864e-01
7.02544689e-01 4.54471648e-01 -4.70587313e-02 -8.85815084e-01
-5.30307829e-01 -5.04807711e-01 -6.21174455e-01 -4.70965564e-01
7.78205812e-01 -8.45848322e-01 -9.23329115e-01 5.71546853e-01
-1.43695176e+00 -1.03569917e-01 -9.05720796e-03 9.00767744e-01
-3.76487106e-01 2.95412362e-01 -7.13387072e-01 -9.02472138e-01
-1.98770240e-01 -1.22581470e+00 7.17943788e-01 4.01011556e-01
-7.09733069e-02 -8.58917594e-01 -1.99776907e-02 -2.59892847e-02
4.31515545e-01 1.95804983e-01 6.30731225e-01 -2.66944408e-01
-9.17671919e-01 -4.01283920e-01 -1.02268577e-01 5.22324294e-02
-1.17529586e-01 1.54756904e-01 -7.91821480e-01 -1.43178841e-02
4.60965425e-01 1.74181506e-01 6.23811483e-01 4.79348481e-01
1.39792550e+00 -2.73796439e-01 -7.23533332e-01 6.68155730e-01
1.31660318e+00 8.89207661e-01 8.11617017e-01 2.94609219e-01
7.95269966e-01 6.04485095e-01 9.64544296e-01 3.91624123e-01
3.37307394e-01 5.53346872e-01 2.54872799e-01 1.40637055e-01
2.62534857e-01 -5.15184760e-01 4.94372040e-01 1.07563436e+00
-3.43686849e-01 -3.02307904e-01 -1.13061285e+00 6.27143562e-01
-2.25087547e+00 -1.43008995e+00 -2.84817815e-01 1.99299848e+00
2.21766517e-01 2.21392319e-01 -3.11096776e-02 1.52174488e-01
7.89289534e-01 2.32995555e-01 -6.93821311e-01 -1.13409109e-01
1.04346976e-01 -5.78234076e-01 3.66777390e-01 1.56729639e-01
-1.11553693e+00 8.54431272e-01 5.98914099e+00 8.50896955e-01
-1.72455168e+00 -7.45920762e-02 4.02621090e-01 -1.82832986e-01
3.76417339e-02 5.13098910e-02 -1.06664073e+00 9.44359958e-01
1.15625966e+00 -3.16559404e-01 3.06437165e-01 1.00062251e+00
5.94364405e-01 4.99478430e-02 -7.73684859e-01 1.18687880e+00
-7.11996853e-02 -1.22522855e+00 -6.45004511e-02 -1.47386551e-01
3.69019091e-01 4.27354395e-01 2.26248905e-01 5.35472512e-01
-2.50504408e-02 -8.81996751e-01 6.91112280e-01 7.63706446e-01
3.89341861e-01 -8.02630603e-01 8.07526112e-01 1.02897000e+00
-1.30410540e+00 -2.07542658e-01 -7.05937624e-01 -4.75106895e-01
3.77206594e-01 5.72732270e-01 -8.15604270e-01 7.22652376e-01
1.02777672e+00 1.07108808e+00 -3.91058356e-01 1.23323631e+00
1.58925101e-01 5.75141966e-01 -5.66326201e-01 -1.88312873e-01
5.09532094e-01 -5.42742908e-01 6.50307059e-01 1.01189971e+00
7.01879561e-01 3.19274515e-01 3.97126734e-01 7.97157407e-01
5.06300390e-01 1.23117622e-02 -1.02575004e+00 -8.08910560e-03
3.88948292e-01 1.00568056e+00 -7.06777751e-01 -2.26031303e-01
-4.51687515e-01 6.96830034e-01 6.46730959e-02 5.26440680e-01
-1.23433936e+00 -3.37239534e-01 2.55141348e-01 1.61455929e-01
2.72226632e-01 -4.78709787e-01 -1.39314502e-01 -1.13016784e+00
1.54068872e-01 -3.09458137e-01 1.50950879e-01 -9.55342054e-01
-1.11875522e+00 7.34504461e-01 7.38093257e-02 -1.81789899e+00
-2.22867519e-01 -2.73479849e-01 -7.13473439e-01 6.05219901e-01
-1.27452028e+00 -1.01522362e+00 -2.98319727e-01 5.29894173e-01
7.19326973e-01 2.65660174e-02 5.95581353e-01 4.97295529e-01
-7.63431549e-01 6.77694380e-02 2.20283791e-01 1.25132561e-01
1.99022576e-01 -5.97218454e-01 1.97601572e-01 9.63900864e-01
-5.70396185e-02 6.23518765e-01 9.35642242e-01 -7.33562887e-01
-1.11654782e+00 -1.44534469e+00 8.40500891e-01 -3.03478777e-01
4.17993546e-01 -1.71956703e-01 -9.26355183e-01 9.08060968e-01
-5.08232564e-02 -4.44646254e-02 3.76844406e-01 -3.08281300e-03
1.45878911e-01 -3.46172065e-01 -6.68691874e-01 7.82881856e-01
1.20530689e+00 -2.96577096e-01 -5.10941505e-01 3.03535104e-01
7.45347798e-01 -5.28277218e-01 -7.47927010e-01 7.08472371e-01
4.48395848e-01 -8.46063316e-01 9.54259396e-01 -3.15919846e-01
4.09883410e-01 -5.13728082e-01 4.75041308e-02 -1.07617199e+00
-3.63433242e-01 -2.69843191e-01 5.74572459e-02 1.10400367e+00
1.73380002e-01 -6.37191057e-01 7.92377472e-01 5.56821525e-01
-1.67864203e-01 -8.22533786e-01 -9.35759962e-01 -9.42491770e-01
-5.17559163e-02 -7.24792242e-01 7.65041053e-01 6.40587449e-01
-3.13017517e-03 4.38982606e-01 -7.11549819e-01 1.88775301e-01
1.52419239e-01 5.08349776e-01 6.85523689e-01 -1.10832214e+00
2.75878936e-01 -2.07883611e-01 -6.15045249e-01 -1.37648618e+00
1.55798689e-01 -7.05466509e-01 1.18456520e-01 -1.33270431e+00
3.34484652e-02 -6.49272203e-01 -2.66394824e-01 8.61955211e-02
-5.48196286e-02 -4.69081074e-01 1.61893144e-01 4.66554314e-01
-8.17583859e-01 8.82098496e-01 1.23691547e+00 -9.06743482e-02
-3.20467085e-01 3.93112540e-01 -8.40625633e-03 1.00987351e+00
8.73483896e-01 -5.02791524e-01 -7.70848393e-01 -6.78313375e-01
2.19253585e-01 4.06941831e-01 2.65963942e-01 -1.34673250e+00
7.62661099e-01 -2.19413325e-01 6.05254054e-01 -1.32396209e+00
6.02900445e-01 -1.12991524e+00 4.53311056e-01 4.65877920e-01
-1.47063389e-01 5.02293050e-01 3.20744932e-01 7.38899529e-01
-3.03134322e-01 -7.85624161e-02 4.64567244e-01 -2.67775357e-01
-1.15877998e+00 6.65181458e-01 -4.56547767e-01 -4.37636048e-01
1.30608904e+00 -5.28494596e-01 -1.70313358e-01 -4.65315431e-01
-4.73315418e-01 5.51437497e-01 3.03814292e-01 7.59012103e-01
7.88080454e-01 -1.25905037e+00 -3.22543055e-01 1.96002036e-01
-1.21996865e-01 1.49821609e-01 6.44493282e-01 7.41846979e-01
-4.64062274e-01 6.92932487e-01 -2.19079569e-01 -7.90820122e-01
-1.09778309e+00 8.60462546e-01 2.80909270e-01 -1.83121026e-01
-9.97485697e-01 5.30938506e-01 1.19254120e-01 -3.62713158e-01
1.49021074e-01 -3.88157099e-01 -5.27768672e-01 -1.40875772e-01
5.92607558e-01 4.84582186e-01 -3.16370219e-01 -8.37310493e-01
-1.14392504e-01 7.39345133e-01 -7.37452507e-02 6.74759671e-02
1.40225661e+00 -4.06479150e-01 1.12200351e-02 1.00550461e+00
1.11438215e+00 -3.38971198e-01 -1.16464400e+00 -7.95087740e-02
-2.84585115e-02 -5.11541188e-01 -2.27820165e-02 -4.04108495e-01
-9.92739856e-01 9.08535182e-01 7.76836812e-01 1.90048009e-01
9.62765992e-01 -5.54174900e-01 1.32712686e+00 4.76668954e-01
6.98148012e-01 -1.10224760e+00 -3.87356192e-01 7.02590108e-01
7.11041510e-01 -1.10566580e+00 -2.50816822e-01 -3.47469956e-01
-7.08666861e-01 1.08015382e+00 7.56276369e-01 -1.03096515e-01
8.19494545e-01 1.96949974e-01 4.28871289e-02 3.44546251e-02
-8.04854691e-01 -7.36465082e-02 -8.57600197e-03 4.15046543e-01
3.36521193e-02 -1.32660046e-01 -8.12423974e-02 5.68022311e-01
-3.93604159e-01 3.79972368e-01 2.58251280e-01 7.08587825e-01
-3.75817031e-01 -8.35576296e-01 -2.64468104e-01 4.93143946e-01
-3.47712904e-01 2.42496818e-01 2.04343393e-01 6.11497939e-01
4.69275653e-01 1.10811448e+00 1.73517838e-01 -6.85854077e-01
2.74466604e-01 -2.25024179e-01 1.34184033e-01 -2.30120853e-01
-2.51110852e-01 -9.51181799e-02 -1.85753807e-01 -7.02543020e-01
-4.49003696e-01 -6.38240814e-01 -1.45032871e+00 -5.63444376e-01
-4.71353650e-01 2.30623007e-01 5.90916932e-01 9.62131917e-01
5.16357243e-01 5.34410775e-01 4.16040868e-01 -7.39519835e-01
-2.20771536e-01 -8.86518240e-01 -3.34322155e-01 4.21671599e-01
1.01289041e-01 -6.31866574e-01 -1.69711977e-01 4.98726740e-02] | [6.322977542877197, 1.0382344722747803] |
47a421e9-ab07-46d9-9f7a-f6fbd091f93b | sentinel-2-sharpening-using-a-single | null | null | https://ieeexplore.ieee.org/document/9464640 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9464640 | Sentinel-2 Sharpening Using a Single Unsupervised Convolutional Neural Network With MTF-Based Degradation Model | The Sentinel-2 (S2) constellation provides multispectral images at 10 m, 20 m, and 60 m resolution bands. Obtaining all bands at 10 m resolution would benefit many applications. Recently, many model-based and deep learning (DL)-based sharpening methods have been proposed. However, the downside of those methods is that the DL-based methods need to be trained separately for the 20 m and the 60 m bands in a supervised manner at reduced resolution, while the model-based methods heavily depend on the hand-crafted image priors. To break the gap, this article proposes a novel unsupervised DL-based S2 sharpening method using a single convolutional neural network (CNN) to sharpen the 20 and 60 m bands at the same time at full resolution. The proposed method replaces the hand-crafted image prior by the deep image prior (DIP) provided by a CNN structure whose parameters are easily optimized using a DL optimizer. We also incorporate the modulation transfer function-based degradation model as a network layer, and add all bands to both network input and output. This setting improves the DIP and exploits the advantage of multitask learning since all S2 bands are highly correlated. Extensive experiments with real S2 data show that our proposed method outperforms competitive methods for reduced-resolution evaluation and yields very high quality sharpened image for full-resolution evaluation. | ['Han V. Nguyen; Magnus O. Ulfarsson; Johannes R. Sveinsson; Mauro Dalla Mura'] | 2021-06-24 | null | null | null | ieee-journal-of-selected-topics-in-applied-7 | ['pansharpening'] | ['computer-vision'] | [ 4.64880794e-01 -2.11312532e-01 -1.29841464e-02 -3.51126701e-01
-1.17797256e+00 -3.83375645e-01 3.09701979e-01 -3.71630669e-01
-5.09980679e-01 6.89371169e-01 3.52456234e-03 -1.17985345e-01
-4.75052655e-01 -9.99318779e-01 -5.90359390e-01 -1.11203897e+00
-3.54741849e-02 -3.20682049e-01 2.75319248e-01 -4.88549381e-01
-1.61230832e-01 6.20047331e-01 -1.38411164e+00 2.08828762e-01
1.15982640e+00 1.25792456e+00 5.29727757e-01 5.72361529e-01
2.09421128e-01 3.92865032e-01 -3.95777524e-01 -1.76006220e-02
7.21033692e-01 -2.23417670e-01 -3.67267728e-01 2.14160532e-01
4.25866842e-01 -3.88434678e-01 -1.61735296e-01 1.23374748e+00
7.38315880e-01 2.48025000e-01 3.21595848e-01 -6.85130119e-01
-4.70778883e-01 1.84045076e-01 -1.18564558e+00 2.13786125e-01
-2.65142918e-01 1.12021439e-01 5.69602430e-01 -7.01636493e-01
3.68416384e-02 9.54001963e-01 1.08642924e+00 1.70749836e-02
-1.28103089e+00 -7.33066022e-01 -1.80235311e-01 1.96149334e-01
-1.55462098e+00 -3.54338616e-01 9.09215212e-01 -3.10615003e-01
6.78318381e-01 6.12161644e-02 3.40106905e-01 5.16179740e-01
1.97320044e-01 2.86692768e-01 1.42321610e+00 -4.23478156e-01
-9.40883532e-02 -1.76060289e-01 -1.75753236e-01 7.30447590e-01
1.50228500e-01 3.11271936e-01 -1.71838403e-01 1.13481686e-01
1.07901192e+00 2.29368545e-02 -6.54613793e-01 9.19848979e-02
-7.55141556e-01 8.99637580e-01 8.31546903e-01 2.55074471e-01
-5.77507019e-01 -1.57998383e-01 3.51383947e-02 2.29275897e-01
7.22856581e-01 3.72768044e-01 -3.59697729e-01 5.10645628e-01
-1.29489732e+00 -3.77997644e-02 -5.09972218e-03 4.84616756e-01
1.11943817e+00 1.93659618e-01 -1.40371293e-01 1.20918357e+00
1.53090477e-01 6.75702929e-01 2.24850953e-01 -9.06261623e-01
4.69092399e-01 2.23052457e-01 3.09184372e-01 -9.33200955e-01
-6.31897926e-01 -7.74921119e-01 -1.23594058e+00 4.66737598e-01
1.87311638e-02 -3.99389118e-01 -9.25808847e-01 1.70438719e+00
1.82430521e-01 1.33634480e-02 1.68324724e-01 1.25977135e+00
6.05187893e-01 8.66386235e-01 -2.72744983e-01 -3.16936344e-01
1.14136529e+00 -1.03085113e+00 -6.20094836e-01 -3.95326406e-01
5.07963225e-02 -8.17652702e-01 7.52569139e-01 3.74934226e-01
-9.16128099e-01 -7.98813462e-01 -1.21503007e+00 2.01370820e-01
-1.22297652e-01 4.41071302e-01 3.88276011e-01 5.75445890e-01
-9.73752618e-01 5.98497272e-01 -6.36000037e-01 -1.70295060e-01
4.01056916e-01 5.94708174e-02 -2.35327482e-01 3.17595825e-02
-1.28269005e+00 9.37313735e-01 4.43085909e-01 4.54540402e-01
-6.95059001e-01 -5.34977913e-01 -8.76276076e-01 1.78135157e-01
3.05412710e-01 -5.10370076e-01 8.67993176e-01 -1.10600626e+00
-1.63819182e+00 7.64006913e-01 6.74192607e-02 -5.30062199e-01
1.68222100e-01 -2.07039967e-01 -6.75000727e-01 5.96016228e-01
2.49490701e-03 4.25641894e-01 1.24787426e+00 -1.49573576e+00
-7.79322565e-01 -1.98383763e-01 9.87964645e-02 2.34490171e-01
-1.92329705e-01 -5.71779720e-02 -3.00050884e-01 -6.59893394e-01
1.95550323e-01 -5.11603773e-01 -1.89055443e-01 5.88949472e-02
-4.23191004e-02 4.80020046e-01 7.24819958e-01 -9.50647652e-01
9.54824805e-01 -2.19600773e+00 -4.27169316e-02 -5.20231575e-03
5.31648695e-02 5.75295687e-01 -4.78324115e-01 2.44092047e-01
-1.44100785e-01 -1.86207578e-01 -6.30167842e-01 -2.52347291e-01
-4.18149024e-01 9.72094238e-02 -1.21756673e-01 5.39880574e-01
2.96541154e-01 5.11105657e-01 -6.00191951e-01 -2.92518526e-01
3.22730958e-01 6.14836752e-01 -2.44634867e-01 2.45366231e-01
1.06003478e-01 3.83477032e-01 -8.70545432e-02 4.38360751e-01
1.40643752e+00 -9.15226564e-02 -1.35799676e-01 -6.35989666e-01
-5.54394305e-01 -2.15710506e-01 -1.18494427e+00 1.47019005e+00
-7.94294655e-01 5.37382424e-01 7.37511337e-01 -1.19890594e+00
1.15849698e+00 2.67881274e-01 3.53518099e-01 -7.43278861e-01
-1.58729330e-01 1.39772967e-01 -3.22833270e-01 -6.43089890e-01
3.92842054e-01 -6.63116992e-01 1.85411260e-01 -9.11843590e-03
-1.66568354e-01 -3.94271672e-01 -3.04899544e-01 -3.56864363e-01
6.00464106e-01 -2.29595974e-02 1.58141047e-01 -1.53277367e-01
6.79390490e-01 -1.93165809e-01 8.23520243e-01 7.61961460e-01
-2.26452742e-02 8.26287270e-01 6.11526780e-02 -3.45110655e-01
-1.00581789e+00 -9.11276340e-01 -2.64250070e-01 7.74711728e-01
4.45117742e-01 2.58764476e-01 -5.80753744e-01 -2.23575756e-01
-1.65078685e-01 4.42977667e-01 -5.50911725e-01 -1.15644731e-01
-3.74023050e-01 -1.08379471e+00 4.28761452e-01 2.53628224e-01
1.39203584e+00 -4.54632580e-01 -5.44691801e-01 1.71983749e-01
-4.22057956e-01 -1.17197371e+00 -2.47673005e-01 2.56601840e-01
-8.72611344e-01 -8.89378548e-01 -6.38342917e-01 -5.70187151e-01
3.76756608e-01 7.01391280e-01 4.86980826e-01 -3.33563328e-01
-1.43053487e-01 1.16019629e-01 -2.59499401e-01 -8.35187733e-02
-1.68715026e-02 -6.35764077e-02 -2.20963240e-01 4.68511134e-01
-2.31814981e-01 -7.32674778e-01 -4.73411083e-01 3.85190517e-01
-1.18690991e+00 3.03098440e-01 9.26334798e-01 1.11779284e+00
6.01665378e-01 6.88792825e-01 4.75950718e-01 -3.32761377e-01
3.30526650e-01 -2.41277099e-01 -9.64887321e-01 2.51334667e-01
-5.71654975e-01 1.49836577e-02 6.55635774e-01 -2.69198775e-01
-1.51310992e+00 -4.73148227e-02 -8.42447579e-02 -4.38954234e-01
-1.50394961e-01 7.44118035e-01 -1.59448594e-01 -4.28245574e-01
7.95926094e-01 3.55523735e-01 8.66891444e-02 -8.07622254e-01
1.46993041e-01 8.28544378e-01 8.79038453e-01 -2.56807059e-01
1.12827241e+00 7.61521339e-01 1.80869177e-01 -1.16080022e+00
-1.22277045e+00 -4.66749728e-01 -3.66169393e-01 -2.58375615e-01
8.61468077e-01 -1.13234019e+00 -2.19443291e-01 6.15562677e-01
-1.06687081e+00 -4.44354117e-01 5.67633426e-03 5.44872940e-01
-3.12788367e-01 6.00109696e-01 -3.89027297e-01 -7.80718684e-01
-4.78484958e-01 -1.09335697e+00 1.08757806e+00 4.59894389e-01
6.83963120e-01 -8.09621096e-01 -4.26404551e-02 5.76833665e-01
8.43300879e-01 4.09365952e-01 7.02303290e-01 2.85093844e-01
-4.51816797e-01 -2.00969934e-01 -8.04492533e-01 6.46576881e-01
3.02058131e-01 -3.57889444e-01 -1.05629098e+00 -5.37633538e-01
3.17724496e-01 -1.48727998e-01 1.24274123e+00 7.79645324e-01
1.20116580e+00 -2.19047248e-01 1.13604560e-01 1.31237519e+00
1.98178101e+00 -8.92266035e-02 7.28231013e-01 5.53451538e-01
2.31281444e-01 3.23164642e-01 4.81793970e-01 4.62755919e-01
7.95543417e-02 7.25765824e-01 8.21844816e-01 -6.64254963e-01
-2.73377240e-01 2.38256052e-01 2.45707646e-01 1.23703413e-01
-4.44789082e-01 7.97319785e-02 -6.37129188e-01 5.08373499e-01
-1.67265844e+00 -1.21609366e+00 -1.02425672e-01 2.18555164e+00
8.93641531e-01 -8.75141844e-02 -1.85058773e-01 -9.42438841e-03
9.83268619e-01 4.82447982e-01 -3.51519883e-01 5.53616509e-02
-4.97348845e-01 3.84257466e-01 1.00246370e+00 9.15108263e-01
-1.25391328e+00 6.17143333e-01 5.82997513e+00 1.02004850e+00
-1.48901093e+00 2.41995081e-01 2.69635409e-01 -7.59928226e-02
-9.74594653e-02 -1.32469475e-01 -5.95618606e-01 2.60652363e-01
5.67825139e-01 1.14901379e-01 5.88043153e-01 5.59482038e-01
6.80722952e-01 -4.38446373e-01 -2.35012189e-01 1.27811348e+00
-1.68322071e-01 -1.17627633e+00 -2.25838855e-01 -1.32610142e-01
8.57667446e-01 1.95944056e-01 1.72430485e-01 7.64612108e-02
-1.36205152e-01 -8.74732435e-01 5.73661983e-01 7.23947763e-01
9.26953971e-01 -9.71549869e-01 9.27515507e-01 2.96457589e-01
-1.19014573e+00 -1.80701450e-01 -5.58736444e-01 -1.04970045e-01
9.78823230e-02 1.06082189e+00 -2.60753006e-01 1.10110319e+00
8.55021536e-01 4.88576233e-01 -3.54452133e-01 1.05439258e+00
-4.40611690e-01 5.56503236e-01 -3.24049473e-01 6.97734058e-01
5.00270724e-01 -4.17012632e-01 6.53652668e-01 1.29090559e+00
4.08103168e-01 2.34385356e-01 2.00938225e-01 7.00825691e-01
1.93126455e-01 -3.20451170e-01 -2.88612008e-01 3.49784970e-01
1.98153719e-01 1.59576857e+00 -3.70336533e-01 -1.48062199e-01
-4.89880711e-01 9.09263968e-01 -6.20120950e-02 5.96945107e-01
-9.16388929e-01 -7.85686731e-01 6.40986323e-01 8.15270171e-02
5.63087642e-01 -3.32442403e-01 -2.58672118e-01 -1.01807153e+00
2.16359403e-02 -8.02427649e-01 2.90052384e-01 -9.72359598e-01
-1.04732800e+00 4.50974703e-01 -2.75748540e-02 -1.27506077e+00
2.86434710e-01 -4.34423804e-01 -5.21288753e-01 1.28601944e+00
-2.27499723e+00 -1.20687437e+00 -7.93073475e-01 5.80725670e-01
3.67030591e-01 -2.10310221e-02 5.26162386e-01 4.43717897e-01
-6.50002122e-01 1.42939597e-01 4.18145478e-01 -7.51983821e-02
8.07036698e-01 -9.33573902e-01 -1.73011482e-01 1.12364042e+00
-6.37531698e-01 6.69187903e-02 6.58632576e-01 -4.56973821e-01
-9.16329205e-01 -1.30404985e+00 5.14081955e-01 4.26104367e-01
4.33958203e-01 2.07060054e-01 -8.99066925e-01 3.95499259e-01
2.47964963e-01 1.44991994e-01 4.66531903e-01 -2.46553972e-01
-1.83440164e-01 -7.27574944e-01 -1.24402249e+00 2.05626100e-01
6.55210555e-01 -5.86971104e-01 -5.31090677e-01 2.15770125e-01
4.43619102e-01 -3.34851265e-01 -9.30873990e-01 6.72612190e-01
4.30176437e-01 -1.25494206e+00 1.06402695e+00 4.77122217e-02
5.53366482e-01 -6.59709811e-01 -3.99500310e-01 -1.44698548e+00
-5.16795754e-01 -5.78076303e-01 2.17221692e-01 9.05792534e-01
3.06921273e-01 -7.70619631e-01 3.24346066e-01 3.89423198e-03
-3.93497050e-01 -4.92712170e-01 -8.70273113e-01 -9.91763949e-01
-1.96448818e-01 -1.82711661e-01 6.23029709e-01 9.53551948e-01
-6.59361064e-01 1.93139195e-01 -5.23286641e-01 9.71170247e-01
1.00823057e+00 5.08464813e-01 5.66632688e-01 -1.06953490e+00
-3.94936800e-01 -5.21494448e-01 3.74694876e-02 -1.01690662e+00
-2.62430996e-01 -5.95210433e-01 1.19934738e-01 -1.53383791e+00
1.14887401e-01 -2.74621159e-01 -1.90849707e-01 5.78909099e-01
-2.62434065e-01 2.48014510e-01 -4.21858653e-02 2.53672481e-01
-1.15884587e-01 7.13624358e-01 1.23044717e+00 -2.94046193e-01
-1.97893709e-01 9.52804554e-03 -5.88375092e-01 7.77415812e-01
7.96538293e-01 -2.58744985e-01 -1.61462843e-01 -8.90266955e-01
1.34317568e-02 2.87620932e-01 5.75524330e-01 -1.18252349e+00
2.13403761e-01 -3.98180604e-01 4.63518023e-01 -5.47059476e-01
4.29509252e-01 -7.49017775e-01 1.90989479e-01 2.26051450e-01
-1.32265761e-01 -5.38945198e-01 3.54865849e-01 2.50848055e-01
-4.09847558e-01 -1.44343734e-01 1.61279857e+00 -1.34984598e-01
-7.30381370e-01 3.09742868e-01 -5.85478246e-02 -3.83117735e-01
6.13951445e-01 -1.46424294e-01 -3.76425207e-01 -3.92493308e-01
-5.89827836e-01 1.62573084e-01 4.27627683e-01 -2.19632108e-02
5.88366508e-01 -1.21697617e+00 -8.32656324e-01 1.24912128e-01
-4.13257815e-02 1.29696816e-01 5.71303785e-01 9.24736679e-01
-3.99419218e-01 1.47913992e-01 -3.75588179e-01 -5.54635465e-01
-8.73388946e-01 2.51694113e-01 7.81572163e-01 -2.82287061e-01
-5.89897990e-01 7.97095358e-01 7.77979270e-02 -3.41322154e-01
-1.85590982e-01 -5.87214753e-02 -1.36748165e-01 1.01194149e-02
5.66381574e-01 3.77535969e-01 1.74115732e-01 -4.67009395e-01
-9.92790312e-02 9.19834256e-01 3.56098890e-01 -7.56711811e-02
1.72498512e+00 -3.10533673e-01 -1.45512670e-01 1.63364504e-02
1.06705415e+00 -5.20603592e-03 -1.64268696e+00 -4.96025085e-01
-4.03754830e-01 -5.73506117e-01 8.34001541e-01 -7.78287709e-01
-1.30675125e+00 9.02806640e-01 7.35847890e-01 1.81837305e-01
1.87040341e+00 -5.51586449e-01 7.15971470e-01 2.94959307e-01
5.95382564e-02 -1.30666828e+00 -3.59806195e-02 3.84922385e-01
9.17693138e-01 -1.31170273e+00 8.46519023e-02 -2.83920109e-01
-3.43436956e-01 1.27362251e+00 3.03399265e-01 1.19053677e-01
3.80675018e-01 1.19369142e-01 8.75544921e-02 -7.47192651e-02
-2.06678629e-01 -5.13618767e-01 4.61417660e-02 6.83103502e-01
2.33436301e-02 -1.56282820e-02 -1.28402576e-01 5.43482780e-01
1.69513762e-01 -2.76761483e-02 4.97404486e-01 6.04380608e-01
-8.40553403e-01 -7.76206851e-01 -9.14153457e-01 2.10043192e-01
-1.33887067e-01 -8.89927596e-02 3.46348405e-01 5.32322466e-01
4.14089322e-01 9.75532293e-01 -1.73605517e-01 -2.45886967e-01
2.90807664e-01 -1.20239310e-01 4.82073963e-01 -2.64650494e-01
-2.00088054e-01 1.61002457e-01 -4.63273339e-02 -4.32629466e-01
-5.73499084e-01 -3.46910775e-01 -8.58399630e-01 -1.67280927e-01
-4.65748161e-01 1.21075861e-01 6.11384153e-01 9.10650253e-01
2.58272648e-01 3.44077259e-01 1.06053400e+00 -1.27958703e+00
-5.89006424e-01 -9.95383024e-01 -9.94588137e-01 -3.20004229e-03
8.66258800e-01 -7.84264684e-01 -5.37211299e-01 1.00087365e-02] | [10.193309783935547, -1.9103509187698364] |
517612e6-2af5-49a6-a9cf-7650cdcbfbe8 | extended-local-binary-patterns-for-efficient | 1907.09160 | null | https://arxiv.org/abs/1907.09160v2 | https://arxiv.org/pdf/1907.09160v2.pdf | Extended Local Binary Patterns for Efficient and Robust Spontaneous Facial Micro-Expression Recognition | Facial Micro-Expressions (MEs) are spontaneous, involuntary facial movements when a person experiences an emotion but deliberately or unconsciously attempts to conceal his or her genuine emotions. Recently, ME recognition has attracted increasing attention due to its potential applications such as clinical diagnosis, business negotiation, interrogations, and security. However, it is expensive to build large scale ME datasets, mainly due to the difficulty of inducing spontaneous MEs. This limits the application of deep learning techniques which require lots of training data. In this paper, we propose a simple, efficient yet robust descriptor called Extended Local Binary Patterns on Three Orthogonal Planes (ELBPTOP) for ME recognition. ELBPTOP consists of three complementary binary descriptors: LBPTOP and two novel ones Radial Difference LBPTOP (RDLBPTOP) and Angular Difference LBPTOP (ADLBPTOP), which explore the local second order information along the radial and angular directions contained in ME video sequences. ELBPTOP is a novel ME descriptor inspired by unique and subtle facial movements. It is computationally efficient and only marginally increases the cost of computing LBPTOP, yet is extremely effective for ME recognition. In addition, by firstly introducing Whitened Principal Component Analysis (WPCA) to ME recognition, we can further obtain more compact and discriminative feature representations, then achieve significantly computational savings. Extensive experimental evaluation on three popular spontaneous ME datasets SMIC, CASME II and SAMM show that our proposed ELBPTOP approach significantly outperforms the previous state-of-the-art on all three single evaluated datasets and achieves promising results on cross-database recognition.Our code will be made available. | ['Matti Pietikäinen', 'Zhong Liu', 'Chengyu Guo', 'Li Liu', 'Jingyun Liang', 'Geng Zhan'] | 2019-07-22 | null | null | null | null | ['micro-expression-recognition'] | ['computer-vision'] | [ 8.99922401e-02 -4.76687759e-01 -3.46989483e-01 -3.52202624e-01
-5.48736453e-01 -2.78401375e-02 5.49584389e-01 -4.62144732e-01
-3.12956989e-01 4.27162439e-01 -2.04354767e-02 2.54198581e-01
-1.95882678e-01 -4.19621974e-01 -3.41697127e-01 -1.29731250e+00
-2.70243734e-01 -1.34791031e-01 -1.29531279e-01 -2.24991292e-01
2.65938044e-01 9.21101332e-01 -1.96872854e+00 3.33214045e-01
3.33128333e-01 1.56187320e+00 -3.20338815e-01 3.84565294e-02
8.94387364e-02 7.62272894e-01 -4.01468009e-01 -4.02112842e-01
2.94395369e-02 -3.61204326e-01 -6.19285107e-01 1.72517464e-01
2.06110358e-01 -2.63586551e-01 -1.78159773e-01 9.25428391e-01
6.68396473e-01 1.76683396e-01 6.52520478e-01 -1.48272491e+00
-4.09722269e-01 -3.11205864e-01 -1.13460982e+00 3.00391316e-01
5.87033093e-01 -1.11403033e-01 4.83070761e-01 -1.13622940e+00
7.08403528e-01 1.26040363e+00 7.54683495e-01 5.32318115e-01
-8.75614882e-01 -1.13566816e+00 -2.02442199e-01 6.68976963e-01
-1.86586916e+00 -6.72416687e-01 9.45596099e-01 -1.72014236e-01
8.91582251e-01 5.12561440e-01 5.18820822e-01 1.27614999e+00
2.46909052e-01 8.72819126e-01 1.43654883e+00 -2.66278058e-01
-3.58936936e-02 2.31166389e-02 -6.95087016e-02 8.86438251e-01
-2.95252055e-01 2.05605347e-02 -7.67660916e-01 -3.10923874e-01
5.01665413e-01 2.70958394e-02 -4.40205902e-01 -3.92107874e-01
-8.75092030e-01 5.84670246e-01 3.87404189e-02 5.13399601e-01
-6.25117481e-01 -2.87779301e-01 4.20223176e-01 1.57649934e-01
2.68030554e-01 -2.98575938e-01 -1.83359787e-01 -5.94671309e-01
-9.29411054e-01 1.06582277e-01 4.83345866e-01 3.51778209e-01
7.12272346e-01 -7.37053603e-02 3.59897427e-02 1.17196631e+00
6.42427877e-02 5.94827235e-01 8.65606964e-01 -7.59905100e-01
3.95089798e-02 4.27817315e-01 -3.90702263e-02 -1.88991427e+00
-5.63934445e-01 1.94375053e-01 -1.14825952e+00 -3.02021243e-02
-4.00668606e-02 2.00843394e-01 -5.51383197e-01 1.64975846e+00
3.69688720e-01 1.69360891e-01 -1.49180919e-01 9.16705310e-01
9.56756592e-01 6.32451296e-01 -1.74141582e-02 -4.28672522e-01
1.31564176e+00 -6.46175265e-01 -9.23993528e-01 1.53275833e-01
2.93362617e-01 -6.75604820e-01 4.84204322e-01 7.77225018e-01
-6.85087740e-01 -3.84519190e-01 -7.91588187e-01 3.03766102e-01
-2.83813238e-01 5.78919768e-01 6.66381180e-01 7.48732507e-01
-8.37704480e-01 4.06874657e-01 -8.83655310e-01 -2.78779447e-01
5.07508755e-01 6.14908934e-01 -9.54686880e-01 1.51194930e-01
-9.80471969e-01 5.63012660e-01 -3.72997224e-02 4.86945212e-01
-2.19165474e-01 -1.83032140e-01 -1.13071644e+00 -5.02934046e-02
1.58989176e-01 2.67857552e-01 7.67318189e-01 -1.07722068e+00
-1.81711686e+00 1.26376915e+00 -5.52772164e-01 2.09122032e-01
1.70412257e-01 8.08746666e-02 -8.74516964e-01 5.05446494e-01
-1.15043543e-01 8.46482098e-01 1.33087337e+00 -8.13208401e-01
-1.92216173e-01 -5.21532416e-01 -6.38222158e-01 -6.60827532e-02
-3.49904716e-01 4.66105700e-01 -2.47768760e-01 -6.89746618e-01
3.30879390e-01 -1.00245190e+00 3.75579119e-01 6.74390942e-02
-8.96922722e-02 -5.03881872e-01 1.21681106e+00 -7.22363234e-01
1.19052088e+00 -2.57777905e+00 9.17904750e-02 4.30861473e-01
-4.94403616e-02 5.86930990e-01 -1.90685764e-01 1.33498132e-01
-4.95324135e-01 -1.11761190e-01 -4.55529010e-03 -9.39353257e-02
3.65009718e-02 9.75675434e-02 -1.00816667e-01 7.96446145e-01
3.20842206e-01 8.79596114e-01 -5.41726112e-01 -5.37640154e-01
2.30288476e-01 6.64676666e-01 -2.89314270e-01 2.03724466e-02
7.14508176e-01 3.28991562e-01 -2.95526117e-01 1.02441609e+00
1.12908924e+00 2.75506619e-02 2.78837122e-02 -3.68919790e-01
1.19542167e-01 -2.31085673e-01 -1.22349882e+00 1.21432698e+00
-1.82863086e-01 6.07435524e-01 2.31964841e-01 -1.40378988e+00
1.16480589e+00 4.13475513e-01 7.36953914e-01 -8.68932962e-01
4.63314861e-01 1.85325965e-01 -4.43618685e-01 -8.13590884e-01
1.83182850e-01 -2.93141216e-01 1.34588763e-05 2.31712699e-01
5.33834845e-02 3.83888245e-01 -4.89945710e-02 -4.68112320e-01
6.70077503e-01 8.86405185e-02 5.82667053e-01 -5.20617738e-02
1.12464523e+00 -6.40889764e-01 8.14560294e-01 9.41154361e-02
-5.98339200e-01 2.00240999e-01 5.16496420e-01 -5.37602484e-01
-2.06217408e-01 -8.82557511e-01 -5.08438945e-01 8.95822227e-01
2.42529839e-01 -2.66940802e-01 -5.79632342e-01 -6.04034603e-01
-8.41972083e-02 1.10462539e-01 -6.94922686e-01 -2.15991452e-01
-5.14311552e-01 -8.84027004e-01 6.69770181e-01 4.33167726e-01
9.16117847e-01 -1.25991511e+00 -6.47722304e-01 -3.77977900e-02
-4.03105855e-01 -1.06959307e+00 -3.69286984e-01 -3.61745566e-01
-4.47343290e-01 -9.82694089e-01 -8.92004490e-01 -8.26945722e-01
5.01208425e-01 2.09903479e-01 4.92352605e-01 -1.81130633e-01
-7.21963584e-01 4.31610107e-01 -3.35468203e-01 -1.44360170e-01
1.36374325e-01 -4.35764819e-01 3.83946806e-01 5.89725971e-01
1.05050063e+00 -5.47903299e-01 -5.36039650e-01 5.83603859e-01
-7.99332976e-01 -3.09590042e-01 7.15684235e-01 9.76949692e-01
6.14345431e-01 -7.53271133e-02 3.09920430e-01 -7.12654665e-02
5.10795236e-01 -3.37902576e-01 -1.61448210e-01 1.80469677e-01
-9.08159092e-02 -2.57517636e-01 1.88262522e-01 -6.12720609e-01
-1.12976182e+00 -1.54804572e-01 -4.01931286e-01 -5.97697139e-01
-3.80387366e-01 3.61081451e-01 -7.04611000e-03 -5.21352530e-01
2.99866498e-01 6.66107059e-01 2.67943114e-01 -3.16300094e-01
-2.26835296e-01 1.05857849e+00 5.09068370e-01 -4.83376116e-01
1.88030317e-01 6.44642711e-01 2.58832395e-01 -1.22559941e+00
-2.96660364e-01 -5.77413380e-01 -4.54045415e-01 -1.77279577e-01
7.28045523e-01 -7.57793784e-01 -1.05149877e+00 8.36391807e-01
-8.35861027e-01 2.80948520e-01 3.18892390e-01 4.08033729e-01
-5.88819802e-01 6.90087855e-01 -5.86016238e-01 -7.89837003e-01
-4.91815507e-01 -1.17512584e+00 1.17032242e+00 9.28366557e-02
-3.87920767e-01 -4.92357224e-01 2.52841244e-04 2.81241179e-01
3.74373764e-01 4.95563447e-01 4.94176269e-01 -2.62015790e-01
-7.48011395e-02 -3.80905420e-01 -3.05333912e-01 4.40357059e-01
3.68571907e-01 -5.02314232e-03 -9.28027570e-01 -1.73707992e-01
4.05509658e-02 -6.39231265e-01 5.90554714e-01 1.57276675e-01
1.55685329e+00 -3.15743715e-01 -3.57945114e-01 7.44089127e-01
9.30291474e-01 3.73315364e-01 8.97956848e-01 3.95704210e-01
2.24755466e-01 6.62391007e-01 7.07972944e-01 6.61689520e-01
1.56280041e-01 8.87096643e-01 6.07729442e-02 3.67955044e-02
4.08115149e-01 2.58956164e-01 5.33776104e-01 5.78743994e-01
-4.78078634e-01 2.57063180e-01 -5.87084174e-01 3.07339221e-01
-1.51648366e+00 -1.34301913e+00 3.66834104e-01 1.82017839e+00
7.12474048e-01 -5.97239256e-01 7.62032298e-03 3.86283249e-01
5.44852257e-01 3.18731427e-01 -2.60307252e-01 -8.19072366e-01
-2.12146521e-01 6.95545435e-01 -1.62126899e-01 -1.07268557e-01
-1.37076938e+00 5.55054784e-01 4.96246529e+00 1.17117310e+00
-1.76772285e+00 1.95463926e-01 7.10661173e-01 -2.17709802e-02
5.39390206e-01 -7.67533600e-01 -6.45990133e-01 4.39351171e-01
7.37572968e-01 1.94504708e-01 2.64704764e-01 1.00973713e+00
7.25949779e-02 -1.40864134e-01 -8.17353964e-01 1.86873376e+00
3.77018481e-01 -1.08795786e+00 -1.24767892e-01 5.66764064e-02
4.12268698e-01 -2.93591619e-01 3.37809771e-01 3.13638061e-01
-8.14553320e-01 -1.13844371e+00 3.66625190e-01 5.26307404e-01
9.80459332e-01 -9.59912598e-01 8.73940527e-01 1.59973025e-01
-1.25011599e+00 -1.83546655e-02 -2.72212893e-01 1.71839282e-01
-2.64921248e-01 1.47699326e-01 -1.98798329e-01 4.35655862e-01
1.07547379e+00 7.69469380e-01 -1.43556163e-01 6.11246288e-01
-5.33209555e-02 1.73543572e-01 -4.41440105e-01 -3.96070592e-02
1.85565189e-01 -1.30091697e-01 4.04164016e-01 1.27388382e+00
5.18203139e-01 3.83869886e-01 -1.81146204e-01 6.47553980e-01
1.04363821e-01 3.79911035e-01 -2.27602288e-01 -1.58357531e-01
1.49034569e-03 1.48703182e+00 -3.61451566e-01 -2.76388731e-02
-3.82741302e-01 1.29506123e+00 1.08613461e-01 1.34694964e-01
-7.79658794e-01 -5.41335821e-01 9.31452036e-01 -1.84704974e-01
5.74031711e-01 -3.75917815e-02 5.68256974e-01 -1.32233799e+00
2.26418763e-01 -1.14004934e+00 2.91621923e-01 -6.14583731e-01
-1.13251257e+00 6.42074227e-01 -2.41779834e-02 -1.30942333e+00
-3.96821469e-01 -8.22879910e-01 -4.20309663e-01 6.31525040e-01
-1.47849751e+00 -1.06740582e+00 -5.45827389e-01 1.01345134e+00
1.02472685e-01 -2.07481757e-01 1.07741976e+00 3.76035482e-01
-7.31613159e-01 1.03854120e+00 -3.99314687e-02 2.42213398e-01
7.18987048e-01 -4.18437570e-01 -4.33309764e-01 3.32553148e-01
-7.69078285e-02 6.74622416e-01 3.53174895e-01 -1.86617315e-01
-1.34231007e+00 -6.55516624e-01 7.46502101e-01 6.99571818e-02
3.71408015e-01 -1.35305971e-01 -8.16065848e-01 4.41017985e-01
-1.16626218e-01 4.83867794e-01 9.62881207e-01 -3.79242867e-01
-3.27446640e-01 -3.70658726e-01 -1.50055921e+00 3.40258151e-01
7.63370395e-01 -4.65658963e-01 -4.30889636e-01 6.84330193e-03
-3.29907179e-01 -4.54654604e-01 -1.06714678e+00 7.26048946e-01
9.52405274e-01 -1.40600383e+00 9.20797288e-01 -1.78698748e-01
1.41029522e-01 8.16891044e-02 -1.02776751e-01 -8.50380063e-01
-1.35926664e-01 -6.30399704e-01 -1.65172189e-01 1.04741740e+00
-3.19555104e-01 -7.56944418e-01 8.43043745e-01 4.13994789e-01
4.11436021e-01 -1.18674219e+00 -1.29172957e+00 -7.93288827e-01
-2.48721316e-01 -3.63533169e-01 6.58056736e-01 1.05454540e+00
4.26717579e-01 -3.94757599e-01 -4.28150266e-01 1.07284375e-02
4.50686216e-01 2.11292073e-01 6.67014301e-01 -8.39974463e-01
1.53302262e-03 -5.78735352e-01 -1.12466836e+00 -8.76978159e-01
3.66334200e-01 -4.92096901e-01 -1.60666019e-01 -5.94771504e-01
1.97212756e-01 -2.54805058e-01 -4.06689405e-01 6.89916074e-01
2.02676162e-01 6.81495130e-01 -7.30144307e-02 1.97273001e-01
-4.49324250e-01 8.20989907e-01 9.36622322e-01 -8.03398192e-02
-1.16952471e-01 -2.44882137e-01 -2.94928551e-01 8.67387772e-01
6.63799584e-01 -2.01967448e-01 6.68605268e-02 3.08249950e-01
-1.44173235e-01 -2.55286135e-02 4.46469009e-01 -9.70852673e-01
-3.09914015e-02 -8.95042792e-02 5.98845899e-01 -5.60061514e-01
6.58078551e-01 -5.55856943e-01 -5.34670055e-03 2.50880957e-01
2.79438376e-01 -1.21578850e-01 3.91161233e-01 2.11881042e-01
-7.45983481e-01 1.48244947e-01 8.77089143e-01 1.35718718e-01
-1.15381956e+00 4.24109757e-01 -5.21800935e-01 -4.38674659e-01
1.22158885e+00 -4.32779580e-01 1.10358754e-02 -3.40140879e-01
-7.57518888e-01 -2.57526547e-01 1.10184990e-01 5.77050507e-01
9.08480525e-01 -1.53558397e+00 -3.75513583e-01 7.37629175e-01
2.39199504e-01 -5.91955841e-01 5.50158024e-01 1.39099991e+00
-4.50990856e-01 6.80356264e-01 -5.05971730e-01 -7.71995664e-01
-1.67707884e+00 4.22871828e-01 2.88716316e-01 2.23206040e-02
-4.72384214e-01 9.59193110e-01 1.71020374e-01 -2.93534458e-01
1.78666309e-01 -1.57638341e-02 -4.01297003e-01 1.08964093e-01
9.41398978e-01 5.29815376e-01 1.93842739e-01 -1.37726092e+00
-7.36376464e-01 9.93900657e-01 -3.93239968e-02 3.61980915e-01
1.27522254e+00 -1.02923959e-01 -5.55775702e-01 -8.87925997e-02
1.63425660e+00 5.54052591e-02 -8.50789368e-01 -1.04293332e-01
-1.53801218e-01 -6.71334147e-01 7.75568783e-02 -3.33294719e-01
-1.15738785e+00 9.04659152e-01 9.11609411e-01 -3.38589996e-01
1.65703785e+00 -1.04527093e-01 9.77836430e-01 3.81561488e-01
5.92301488e-01 -9.03838754e-01 2.05987617e-01 1.55238584e-01
1.02275050e+00 -1.16744137e+00 -1.03972048e-01 -4.27660346e-01
-7.35795021e-01 1.37935507e+00 4.48497146e-01 3.44158858e-02
9.06740189e-01 1.55558893e-02 1.14296727e-01 -2.63996094e-01
-3.05116385e-01 2.29013979e-01 5.97136676e-01 4.69074339e-01
3.00890744e-01 -1.61544159e-01 -5.68466544e-01 7.18689680e-01
-8.93605351e-02 3.33032191e-01 -9.58561245e-03 1.09291351e+00
-2.23837748e-01 -9.79509532e-01 -4.90185827e-01 2.67274797e-01
-7.55778253e-01 3.60784411e-01 -1.30738258e-01 8.62637341e-01
4.71789688e-01 6.86083972e-01 1.80296808e-01 -4.36981052e-01
5.13174161e-02 1.31550923e-01 6.42512619e-01 5.34170419e-02
1.40554637e-01 7.51611739e-02 -1.21185482e-01 -1.09154987e+00
-9.11517322e-01 -8.55464697e-01 -1.20832598e+00 -4.40873832e-01
-3.07346340e-02 -5.51353619e-02 4.90165859e-01 9.39654112e-01
4.82995480e-01 -2.46060714e-01 6.97607100e-01 -1.13074338e+00
-3.71686637e-01 -7.87441373e-01 -8.30073237e-01 5.84920645e-01
3.58971149e-01 -1.16992152e+00 -3.79709750e-01 -1.58105493e-01] | [13.633965492248535, 1.7720701694488525] |
ca6a21f6-0211-4256-bd85-0d66b5f78d33 | sketchmate-deep-hashing-for-million-scale | 1804.01401 | null | http://arxiv.org/abs/1804.01401v1 | http://arxiv.org/pdf/1804.01401v1.pdf | SketchMate: Deep Hashing for Million-Scale Human Sketch Retrieval | We propose a deep hashing framework for sketch retrieval that, for the first
time, works on a multi-million scale human sketch dataset. Leveraging on this
large dataset, we explore a few sketch-specific traits that were otherwise
under-studied in prior literature. Instead of following the conventional sketch
recognition task, we introduce the novel problem of sketch hashing retrieval
which is not only more challenging, but also offers a better testbed for
large-scale sketch analysis, since: (i) more fine-grained sketch feature
learning is required to accommodate the large variations in style and
abstraction, and (ii) a compact binary code needs to be learned at the same
time to enable efficient retrieval. Key to our network design is the embedding
of unique characteristics of human sketch, where (i) a two-branch CNN-RNN
architecture is adapted to explore the temporal ordering of strokes, and (ii) a
novel hashing loss is specifically designed to accommodate both the temporal
and abstract traits of sketches. By working with a 3.8M sketch dataset, we show
that state-of-the-art hashing models specifically engineered for static images
fail to perform well on temporal sketch data. Our network on the other hand not
only offers the best retrieval performance on various code sizes, but also
yields the best generalization performance under a zero-shot setting and when
re-purposed for sketch recognition. Such superior performances effectively
demonstrate the benefit of our sketch-specific design. | ['Yi-Zhe Song', 'Kaiyue Pang', 'Zhanyu Ma', 'Tongtong Yuan', 'Timothy M. Hospedales', 'Tao Xiang', 'Peng Xu', 'Yongye Huang', 'Jun Guo'] | 2018-04-04 | sketchmate-deep-hashing-for-million-scale-1 | http://openaccess.thecvf.com/content_cvpr_2018/html/Xu_SketchMate_Deep_Hashing_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Xu_SketchMate_Deep_Hashing_CVPR_2018_paper.pdf | cvpr-2018-6 | ['sketch-recognition'] | ['computer-vision'] | [-1.06161386e-01 -5.40782392e-01 -3.97787124e-01 -1.56581059e-01
-7.16000736e-01 -6.04872465e-01 6.64374113e-01 -9.84539315e-02
-2.27248460e-01 2.62044102e-01 2.72981733e-01 -1.05485678e-01
-7.18990192e-02 -7.70241797e-01 -5.87163627e-01 -5.29851735e-01
-2.71944612e-01 4.80331481e-01 1.63800225e-01 -3.81999761e-01
2.91271180e-01 7.39613652e-01 -1.44374001e+00 1.41606122e-01
2.34932706e-01 1.19768465e+00 1.14099637e-01 6.23551011e-01
-1.46339893e-01 2.99750596e-01 -3.44800234e-01 -4.75563586e-01
4.56697136e-01 -1.12733454e-01 -4.46306676e-01 -1.41068503e-01
8.97771776e-01 -9.14755285e-01 -1.02176106e+00 5.25833189e-01
6.62362933e-01 -4.57670353e-02 6.14544868e-01 -1.24564898e+00
-1.01517844e+00 4.09179270e-01 -3.23394537e-01 -1.94940552e-01
2.78845668e-01 4.24842268e-01 1.40270948e+00 -1.05129135e+00
6.08126283e-01 1.22617829e+00 6.94163561e-01 6.83476269e-01
-1.20947397e+00 -9.18020129e-01 4.73168939e-02 -6.49263933e-02
-1.58994627e+00 -3.06398273e-01 8.86432111e-01 -4.22512777e-02
8.08903396e-01 -4.37169224e-02 8.32728267e-01 1.47934616e+00
-6.09820150e-03 1.00301337e+00 5.31090140e-01 -3.38955075e-02
1.94457665e-01 -8.23118761e-02 -1.65343791e-01 8.20391655e-01
-3.65432985e-02 1.81011427e-02 -6.02151334e-01 -2.15548977e-01
1.09692073e+00 4.93927211e-01 -4.76998314e-02 -6.33040309e-01
-1.27144790e+00 8.37603271e-01 5.73711753e-01 2.47120962e-01
-1.28536493e-01 6.41657352e-01 6.52262032e-01 4.51045543e-01
-1.22885883e-01 3.75118971e-01 -3.04491282e-01 -2.41426915e-01
-1.18865848e+00 5.10244608e-01 8.50167155e-01 1.17976820e+00
7.18045890e-01 2.04541478e-02 -3.36631238e-01 8.36722434e-01
7.60589316e-02 6.12424254e-01 2.92513460e-01 -7.25327194e-01
5.20320117e-01 5.12871444e-01 -1.62548676e-01 -1.00883830e+00
-2.82537919e-02 -1.27886876e-01 -1.05679059e+00 -1.47344053e-01
3.36829424e-01 4.92395133e-01 -7.67780185e-01 1.95660067e+00
-1.68309376e-01 -6.27418235e-02 -3.14534605e-01 8.17120552e-01
6.15281522e-01 5.44382393e-01 -4.14325744e-02 4.91424978e-01
1.45588160e+00 -9.36809301e-01 -2.98931539e-01 9.55123603e-02
1.27345741e-01 -5.18227756e-01 1.39007652e+00 3.02329987e-01
-1.15355372e+00 -6.50718272e-01 -1.31702232e+00 -4.66782361e-01
-6.21298552e-01 -3.85139696e-02 8.93477201e-01 4.73072171e-01
-1.10188735e+00 6.53644443e-01 -4.72885132e-01 -5.66702902e-01
5.03790259e-01 2.85224259e-01 -4.70625520e-01 -2.84760684e-01
-1.13620126e+00 4.47361737e-01 -1.14710726e-01 -3.23694237e-02
-8.24927032e-01 -8.03339779e-01 -8.80581260e-01 5.46827495e-01
1.85088754e-01 -5.77115297e-01 9.15005922e-01 -4.04869318e-01
-1.51226890e+00 5.84651172e-01 7.98445642e-02 -2.07190111e-01
4.62807089e-01 -8.27142671e-02 -1.43580988e-01 4.68788803e-01
-1.44144282e-01 1.06637061e+00 1.13070107e+00 -9.33511615e-01
-2.32784048e-01 -2.44547576e-01 -3.93794179e-02 -1.81513205e-01
-8.85397136e-01 -2.77852952e-01 -8.78207743e-01 -1.07988477e+00
-2.49634415e-01 -1.01136398e+00 8.17260444e-02 7.06234932e-01
-9.21463072e-02 -3.08597833e-01 9.72053468e-01 -2.99804211e-01
1.44618571e+00 -2.24003482e+00 2.52510309e-01 2.57024080e-01
2.18294755e-01 3.25971693e-01 -4.28767979e-01 9.53694642e-01
2.80167192e-01 1.55318275e-01 -3.17120582e-01 -5.50746322e-01
4.47781473e-01 2.41873488e-01 -9.11367953e-01 2.38550618e-01
3.23926598e-01 1.30709529e+00 -7.09035695e-01 -4.47593361e-01
-2.45698285e-03 4.41542476e-01 -7.54537761e-01 3.35021049e-01
-1.51367471e-01 -1.04772538e-01 -4.29373980e-01 9.36779559e-01
5.48223078e-01 -3.82540345e-01 1.71254426e-01 -2.01889262e-01
1.24233335e-01 1.00115329e-01 -1.04408491e+00 2.14015794e+00
-4.76494014e-01 5.75886965e-01 -3.87909599e-02 -7.20749915e-01
8.78824234e-01 2.76261061e-01 3.61505657e-01 -8.39521289e-01
-3.45190257e-01 3.66193473e-01 -4.59045291e-01 -1.08993828e-01
7.11689830e-01 -1.60893157e-01 -3.95182490e-01 7.52711892e-01
1.99466109e-01 -1.32128641e-01 -9.64672044e-02 4.75197375e-01
1.21919024e+00 -4.55247350e-02 6.65798113e-02 -1.70975804e-01
2.11140975e-01 -6.63821399e-01 2.39083990e-01 9.85723674e-01
-9.01975930e-02 7.75081694e-01 5.13729095e-01 -7.64651477e-01
-1.44419241e+00 -1.16353416e+00 -9.63635147e-02 1.24692941e+00
2.15864927e-01 -5.62307119e-01 -1.30006760e-01 -5.96393585e-01
6.47967994e-01 -6.43501505e-02 -6.89848363e-01 -1.98437423e-01
-6.51010633e-01 -1.14656724e-02 9.51114297e-01 7.14150667e-01
2.65685558e-01 -1.18178117e+00 -5.14055431e-01 1.85452148e-01
4.18958664e-01 -9.89242196e-01 -9.64442313e-01 1.50709882e-01
-6.77021325e-01 -6.29796028e-01 -1.03678310e+00 -7.56034434e-01
2.81312019e-01 4.76032704e-01 1.09116709e+00 6.31464839e-01
-4.52613503e-01 7.40195215e-01 -1.90969691e-01 1.17026418e-01
4.17926721e-02 4.50800687e-01 -5.54418890e-03 -2.96296719e-02
-4.32136320e-02 -9.51912463e-01 -8.02119851e-01 3.07622850e-01
-1.21291029e+00 -2.56924391e-01 7.84358084e-01 1.20553327e+00
6.27628937e-02 -2.90399700e-01 6.56984866e-01 -2.93027163e-01
4.86955673e-01 -2.61092335e-01 -4.19030368e-01 4.50353205e-01
-5.53668797e-01 3.35911959e-01 8.03874552e-01 -7.32796788e-01
-4.65578467e-01 -2.12985486e-01 -1.15720272e-01 -6.87498331e-01
1.42616272e-01 9.73409936e-02 -4.00214270e-02 -2.28281528e-01
2.48917311e-01 3.50161880e-01 1.20677449e-01 -5.10891855e-01
5.08007824e-01 4.49118584e-01 5.04545927e-01 -1.16660798e+00
9.80608225e-01 6.16403699e-01 1.44772246e-01 -7.82297552e-01
-2.51336217e-01 -3.18266392e-01 -5.67407966e-01 2.18169138e-01
4.80136424e-01 -8.55825841e-01 -1.05216217e+00 3.35929006e-01
-1.06590247e+00 -4.19143558e-01 -9.20305252e-02 -1.02097355e-01
-5.68332016e-01 6.62658274e-01 -8.93033028e-01 -6.92165315e-01
-6.33635342e-01 -1.18336439e+00 1.50652409e+00 -7.44937733e-02
-1.22553192e-01 -6.99984729e-01 2.45715268e-02 -1.63589969e-01
7.38034189e-01 -1.84254587e-01 1.26327252e+00 -3.45229357e-01
-9.86877143e-01 -2.56029874e-01 -6.01899981e-01 1.12128295e-01
-1.03983998e-01 2.11476791e-03 -7.89063871e-01 -5.74776232e-01
-6.52156949e-01 -8.26828241e-01 1.10419679e+00 -1.99358463e-01
1.33331084e+00 -2.97120035e-01 -9.63174850e-02 8.12495649e-01
1.33848047e+00 -1.61858708e-01 6.06038511e-01 4.92004789e-02
7.38847554e-01 2.39586115e-01 1.30604446e-01 6.14274263e-01
5.28687477e-01 1.01323640e+00 1.99807152e-01 -1.32944193e-02
-1.66754201e-01 -7.21019745e-01 1.60995856e-01 7.50433624e-01
3.78466159e-01 -9.23719406e-02 -6.78955317e-01 6.61047101e-01
-1.84937596e+00 -9.56983626e-01 8.22973967e-01 2.10242581e+00
8.82356524e-01 1.05926529e-01 3.82297128e-01 -9.39828902e-02
2.68890858e-01 4.57661957e-01 -6.29811406e-01 -3.34673733e-01
-1.05421230e-01 5.85535765e-01 2.06738412e-01 1.51954601e-02
-9.05741394e-01 9.92607951e-01 6.16943169e+00 9.04357076e-01
-1.32004762e+00 -3.08124095e-01 2.27247566e-01 -1.30830407e-01
-5.29877365e-01 -5.03928866e-03 -7.02687800e-01 5.80690861e-01
4.78197753e-01 3.11067939e-01 8.37774098e-01 7.67658770e-01
-6.09581470e-01 3.82984728e-01 -1.44128311e+00 1.11418676e+00
-2.17784569e-03 -1.42848897e+00 5.15479624e-01 2.11680412e-01
3.50187331e-01 -3.03777069e-01 3.87601405e-01 6.14902377e-01
-1.32235765e-01 -1.11749053e+00 8.63968253e-01 3.63473535e-01
1.15674913e+00 -6.47359729e-01 1.98715493e-01 2.85085496e-02
-1.59449589e+00 -2.76967853e-01 -5.40790379e-01 2.84648184e-02
-2.40505068e-03 1.53661847e-01 -4.44984764e-01 3.11257482e-01
5.82858920e-01 8.35436940e-01 -5.84582031e-01 7.62645781e-01
4.44820635e-02 8.95557329e-02 -2.87922919e-01 -1.22724816e-01
4.93928164e-01 5.12705259e-02 2.25450993e-01 1.45221651e+00
3.21940958e-01 -1.47547200e-01 1.07955001e-01 9.80662644e-01
-4.54716593e-01 -1.48156390e-01 -9.19068515e-01 -2.51882821e-01
8.75039697e-01 1.23246944e+00 -4.74749476e-01 -3.02976370e-01
-3.39880258e-01 1.13930845e+00 6.46297693e-01 4.10502106e-01
-6.74601436e-01 -5.35734594e-01 9.30846930e-01 1.21663767e-03
8.90505672e-01 -5.08256853e-01 -1.88948229e-01 -1.24056637e+00
1.91784844e-01 -8.21057320e-01 2.83552736e-01 -4.64436054e-01
-1.58561110e+00 3.79052490e-01 -2.82266319e-01 -9.82662439e-01
-9.42616090e-02 -6.96256697e-01 -5.99718750e-01 7.14392483e-01
-1.46217346e+00 -1.48122191e+00 -1.64299875e-01 5.56115031e-01
4.97651786e-01 -3.24595481e-01 9.55968261e-01 6.13622367e-01
-5.25943637e-01 1.21707428e+00 -9.23165008e-02 2.98409879e-01
1.00729799e+00 -1.19212151e+00 8.91067326e-01 4.07670259e-01
2.92143404e-01 1.14737856e+00 1.26980215e-01 -4.51917917e-01
-2.14145184e+00 -7.79046535e-01 6.75495267e-01 -4.79122788e-01
7.69471765e-01 -9.67660069e-01 -9.20504332e-01 3.18344444e-01
-6.29440323e-02 2.85523266e-01 4.35753495e-01 1.66904062e-01
-1.22343647e+00 -3.27017903e-01 -9.04789746e-01 7.75214970e-01
1.14535666e+00 -1.20480812e+00 -3.22786540e-01 -5.27014323e-02
7.61629164e-01 -1.23067088e-01 -1.11289096e+00 2.16176763e-01
1.48662007e+00 -8.04207623e-01 1.42320669e+00 -7.61815667e-01
5.79568088e-01 -2.50771735e-02 -4.57310051e-01 -6.14278793e-01
-4.22933728e-01 -7.30873525e-01 -3.69614631e-01 1.15155375e+00
1.26357213e-01 -3.02310109e-01 7.50819266e-01 5.61134756e-01
2.62008518e-01 -1.26732290e+00 -7.97341585e-01 -9.10094798e-01
1.07258029e-01 -2.32255831e-01 8.26244116e-01 7.21088886e-01
-1.47942305e-01 2.13113889e-01 -7.81202435e-01 -1.97887331e-01
5.78047633e-01 5.40342331e-01 1.02694309e+00 -1.15304577e+00
-3.30890983e-01 -7.53584743e-01 -3.82826656e-01 -1.53468263e+00
2.73879230e-01 -7.24350989e-01 -1.20398872e-01 -9.70342457e-01
3.84232670e-01 -7.84142733e-01 -3.36751521e-01 5.24448752e-01
-7.62683824e-02 4.94512409e-01 6.10330701e-01 3.16364408e-01
-7.34197319e-01 8.85387123e-01 1.03628385e+00 -3.06872070e-01
1.12530962e-01 -3.52931023e-01 -6.92565024e-01 1.23586673e-02
3.29057932e-01 -1.80972651e-01 -3.41806501e-01 -4.37048852e-01
2.31770575e-01 2.34751225e-01 6.88509583e-01 -8.27535510e-01
5.02500117e-01 1.43707335e-01 3.66537660e-01 -6.40689790e-01
5.95584989e-01 -8.19048464e-01 -1.18191563e-01 3.21954787e-01
-4.00850296e-01 2.73880541e-01 1.45269617e-01 7.58394122e-01
-3.48150581e-02 7.51743391e-02 7.07409978e-01 -7.37947524e-02
-5.46566546e-01 7.65059829e-01 1.05229944e-01 -9.99409854e-02
3.83736849e-01 -2.23142728e-01 -2.49803260e-01 -5.17398894e-01
-9.19785909e-03 7.16392770e-02 8.11039865e-01 6.71483159e-01
7.06652641e-01 -1.80928826e+00 -3.52383733e-01 3.47630560e-01
4.23140049e-01 -2.99375266e-01 1.72606215e-01 3.48266602e-01
-2.61611521e-01 4.91918147e-01 -4.13864076e-01 -5.22248805e-01
-6.87966049e-01 8.67665052e-01 -1.25524774e-01 -3.17236990e-01
-9.56289172e-01 6.54413879e-01 3.02282602e-01 -3.54188204e-01
6.71229541e-01 -2.80789375e-01 5.06352365e-01 6.79349303e-02
5.24714231e-01 1.52454436e-01 -5.98107167e-02 -2.02592254e-01
-2.35741287e-01 7.97525883e-01 -2.41119802e-01 -9.43342373e-02
1.37759638e+00 1.95804685e-01 4.57117371e-02 2.78087169e-01
1.61992443e+00 -8.40272158e-02 -1.39243817e+00 -3.49729985e-01
-4.44634818e-02 -4.25961852e-01 -2.23639503e-01 -5.47905922e-01
-9.69403982e-01 1.04767227e+00 2.05777615e-01 6.34373445e-03
8.83978903e-01 -1.50888845e-01 1.32614708e+00 7.43906260e-01
5.38892686e-01 -8.95030856e-01 5.90308726e-01 5.94967425e-01
9.53337789e-01 -1.13868558e+00 4.89903726e-02 3.00156772e-01
-5.47661006e-01 1.34098768e+00 3.19902301e-01 -2.46252999e-01
5.08806825e-01 1.98383600e-01 -3.99032027e-01 -3.52682889e-01
-8.25697422e-01 1.52844623e-01 3.60374779e-01 1.61031082e-01
1.64518669e-01 -9.11165252e-02 1.99478760e-01 5.06051898e-01
5.71475066e-02 -1.67468429e-01 -6.77187964e-02 8.62260699e-01
-1.74144968e-01 -1.18979645e+00 -4.39240374e-02 3.49928975e-01
6.84228316e-02 -1.03279971e-01 -3.65749568e-01 8.81493628e-01
-2.58457184e-01 2.17751443e-01 -4.07365933e-02 -4.08197671e-01
2.96591580e-01 7.11342692e-02 6.57998741e-01 -3.21107358e-01
-6.40907228e-01 -3.05436820e-01 -4.55912173e-01 -7.10292518e-01
1.26532629e-01 -2.99522936e-01 -7.64622033e-01 -5.39260626e-01
1.67813301e-01 -2.11583272e-01 5.98233640e-01 5.23586035e-01
5.37101090e-01 9.09127444e-02 6.87122345e-01 -1.25868618e+00
-8.95155847e-01 -6.65594697e-01 -7.14214027e-01 4.83416736e-01
5.81423819e-01 -8.03926110e-01 -2.05348745e-01 -5.13998210e-01] | [11.69491958618164, 0.5570034980773926] |
c71eefbb-2eea-4d0d-b0c8-cec0081cd323 | class-aware-contrastive-semi-supervised | 2203.02261 | null | https://arxiv.org/abs/2203.02261v3 | https://arxiv.org/pdf/2203.02261v3.pdf | Class-Aware Contrastive Semi-Supervised Learning | Pseudo-label-based semi-supervised learning (SSL) has achieved great success on raw data utilization. However, its training procedure suffers from confirmation bias due to the noise contained in self-generated artificial labels. Moreover, the model's judgment becomes noisier in real-world applications with extensive out-of-distribution data. To address this issue, we propose a general method named Class-aware Contrastive Semi-Supervised Learning (CCSSL), which is a drop-in helper to improve the pseudo-label quality and enhance the model's robustness in the real-world setting. Rather than treating real-world data as a union set, our method separately handles reliable in-distribution data with class-wise clustering for blending into downstream tasks and noisy out-of-distribution data with image-wise contrastive for better generalization. Furthermore, by applying target re-weighting, we successfully emphasize clean label learning and simultaneously reduce noisy label learning. Despite its simplicity, our proposed CCSSL has significant performance improvements over the state-of-the-art SSL methods on the standard datasets CIFAR100 and STL10. On the real-world dataset Semi-iNat 2021, we improve FixMatch by 9.80% and CoMatch by 3.18%. Code is available https://github.com/TencentYoutuResearch/Classification-SemiCLS. | ['Long Zeng', 'Chengjie Wang', 'Wei zhang', 'Feng Zheng', 'Yong liu', 'Guannan Jiang', 'Shuyi Zhang', 'Kai Wu', 'Fan Yang'] | 2022-03-04 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Yang_Class-Aware_Contrastive_Semi-Supervised_Learning_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Yang_Class-Aware_Contrastive_Semi-Supervised_Learning_CVPR_2022_paper.pdf | cvpr-2022-1 | ['semi-supervised-image-classification'] | ['computer-vision'] | [ 2.59283572e-01 -1.04598656e-01 -4.26843524e-01 -8.43680322e-01
-1.41630912e+00 -6.35645866e-01 5.05397975e-01 1.71705902e-01
-5.06505907e-01 9.34592128e-01 -2.09130347e-01 -1.77278608e-01
-7.05348849e-02 -3.22943479e-01 -6.60626292e-01 -1.14908278e+00
2.59404689e-01 3.82985830e-01 1.73294097e-01 1.90229580e-01
-1.12103224e-01 1.30360082e-01 -1.51509607e+00 4.15310651e-01
1.12934029e+00 1.04810059e+00 1.53642684e-01 1.92461178e-01
-1.33470625e-01 1.01027858e+00 -6.20599210e-01 -2.23243043e-01
2.45670035e-01 -4.63259608e-01 -6.12914205e-01 3.00569177e-01
5.61808586e-01 9.75881293e-02 -3.22332233e-02 1.36274111e+00
7.79510498e-01 1.57876667e-02 8.19321811e-01 -1.50739968e+00
-6.01270735e-01 6.73194468e-01 -9.11041915e-01 -7.16634020e-02
-2.60113001e-01 1.37638837e-01 9.26972508e-01 -1.05598629e+00
3.87664169e-01 1.19487596e+00 7.22452521e-01 5.89739859e-01
-1.37748694e+00 -1.22747374e+00 1.42895669e-01 2.13143915e-01
-1.51585174e+00 -4.40217823e-01 7.36422777e-01 -4.11488503e-01
2.27601752e-01 1.89834505e-01 -6.72368845e-03 1.14277446e+00
-3.14037412e-01 1.12135518e+00 1.70508194e+00 -4.48260367e-01
3.95743787e-01 3.20373744e-01 4.05392259e-01 4.92572963e-01
2.51387265e-02 1.61202192e-01 -4.00868505e-01 -8.97513181e-02
1.20897457e-01 -5.36900833e-02 -2.58360595e-01 -1.72815010e-01
-1.15006804e+00 6.44876719e-01 4.86269414e-01 2.78851185e-02
2.35550273e-02 -1.30925640e-01 4.49702978e-01 1.54358596e-01
1.02670181e+00 1.12922400e-01 -7.06030011e-01 1.88261792e-01
-1.21484506e+00 1.01938462e-02 3.11355561e-01 1.01636183e+00
7.10054696e-01 -6.61072433e-02 -3.50992620e-01 1.30741501e+00
3.86575401e-01 7.90037453e-01 4.50684398e-01 -1.06506932e+00
4.66388464e-01 4.24892038e-01 -1.20910458e-01 -4.74534422e-01
-5.01029968e-01 -1.06327844e+00 -9.72664118e-01 1.80611476e-01
6.36554062e-01 -1.19819202e-01 -1.12343919e+00 1.88073862e+00
4.45741355e-01 4.09411073e-01 -6.92338347e-02 8.17614913e-01
9.24581826e-01 5.24818838e-01 4.09077048e-01 -3.96678001e-01
1.14980876e+00 -1.16029584e+00 -7.70439625e-01 -1.96972221e-01
1.09830451e+00 -6.54696167e-01 1.18654346e+00 5.05884588e-01
-4.64883685e-01 -6.37941599e-01 -9.85081255e-01 1.53269917e-01
-2.77915865e-01 3.43225271e-01 3.21280122e-01 7.08977103e-01
-7.74617434e-01 3.74056071e-01 -4.41974550e-01 -7.78260082e-02
8.93437266e-01 8.22300464e-02 -3.60580862e-01 -4.84377235e-01
-1.23006153e+00 4.45455849e-01 5.56543231e-01 -1.77480280e-01
-9.03019190e-01 -8.49222958e-01 -6.35445833e-01 -1.97183773e-01
6.44782066e-01 -9.22477171e-02 1.33232343e+00 -9.35242891e-01
-9.79171038e-01 1.15417075e+00 -8.14772919e-02 -2.17994720e-01
6.06190860e-01 6.24269359e-02 -5.19997120e-01 -7.71398246e-02
3.35059255e-01 7.11378276e-01 6.88374639e-01 -1.65712523e+00
-8.91247272e-01 -3.97349834e-01 -3.41475815e-01 2.32168108e-01
-3.82592350e-01 -1.59723222e-01 -3.40654552e-01 -9.26988423e-01
2.37223178e-01 -1.09319067e+00 -8.44054520e-02 1.88999884e-02
-5.17015100e-01 -4.14070487e-01 7.19526231e-01 -2.76087314e-01
1.25537848e+00 -2.35338306e+00 -4.22820747e-01 -9.00768209e-03
3.36185873e-01 5.55201888e-01 -2.74408311e-01 7.47860968e-02
-4.07061547e-01 1.64935127e-01 -4.41351324e-01 -7.03657925e-01
-1.01750337e-01 2.11606368e-01 -2.09327921e-01 5.72650790e-01
1.39532462e-01 6.77873552e-01 -1.21166646e+00 -7.68757224e-01
1.33800460e-03 2.69265354e-01 -1.84123635e-01 9.82905403e-02
-4.51023951e-02 6.67754054e-01 -2.95843214e-01 8.35386395e-01
8.67373884e-01 -4.26241666e-01 -5.56729995e-02 -1.95097014e-01
1.53691247e-01 5.41162156e-02 -1.09527457e+00 1.69011939e+00
-3.77145320e-01 3.54888797e-01 -1.83970362e-01 -9.99384105e-01
8.03002894e-01 2.21806511e-01 4.61531967e-01 -7.06813812e-01
1.45427808e-01 3.99097711e-01 -1.81217238e-01 -3.37392241e-01
-5.97895831e-02 -2.55705327e-01 6.93826973e-02 4.88586009e-01
1.73276737e-01 6.94704875e-02 2.55767941e-01 3.39482427e-01
9.22841549e-01 3.31744626e-02 1.16511181e-01 -3.85648608e-01
3.58074278e-01 -1.61437184e-01 8.30695093e-01 7.76294351e-01
-5.83489656e-01 8.23111892e-01 2.97240347e-01 -8.19164235e-03
-6.59175038e-01 -1.12599373e+00 -5.55332839e-01 1.21919537e+00
-5.50411381e-02 -1.53483182e-01 -7.44687378e-01 -1.12819362e+00
6.26129657e-02 8.64789605e-01 -6.00447297e-01 -3.15800846e-01
-6.28822520e-02 -1.26003611e+00 6.13738358e-01 2.49754742e-01
7.49508440e-01 -8.14556122e-01 1.87311769e-01 1.11126341e-02
-3.80542248e-01 -1.18552077e+00 -5.59571862e-01 4.92296219e-01
-6.37022316e-01 -1.03440344e+00 -7.47369528e-01 -8.03230286e-01
7.65216172e-01 5.33903837e-01 1.10845697e+00 -9.42779705e-02
-7.55386055e-02 -2.14578927e-01 -6.43902123e-01 -3.62018287e-01
-4.62708473e-01 7.76689276e-02 1.17701992e-01 2.92071462e-01
4.59231287e-01 -3.35896134e-01 -4.62935925e-01 5.54203093e-01
-7.72094667e-01 -4.98513095e-02 4.83884901e-01 1.11145186e+00
8.66268754e-01 4.57926095e-01 9.23364520e-01 -1.46033037e+00
2.06413180e-01 -7.61158526e-01 -3.43301922e-01 3.23293746e-01
-1.09557462e+00 -9.68839005e-02 5.91044664e-01 -5.79670012e-01
-1.17214298e+00 3.59980054e-02 -3.53144370e-02 -4.69658405e-01
-3.31941038e-01 2.79594243e-01 -3.02925646e-01 1.86236501e-01
9.12504613e-01 1.03411935e-01 -1.30606219e-01 -6.18469357e-01
3.18693221e-01 1.15738559e+00 4.32197601e-01 -6.66254580e-01
8.30046237e-01 4.18433160e-01 -1.46147132e-01 -4.17066425e-01
-1.62660217e+00 -7.21393824e-01 -5.26352644e-01 -1.96549267e-01
4.79739308e-01 -1.35055041e+00 -2.68822223e-01 9.40549016e-01
-6.28015935e-01 -6.48210526e-01 -3.22649449e-01 3.55366796e-01
-2.32615083e-01 2.31604382e-01 -5.65163970e-01 -8.69541526e-01
-1.70975804e-01 -1.09723127e+00 1.08103549e+00 1.21810913e-01
2.12384939e-01 -8.48297954e-01 -1.12909272e-01 5.82015634e-01
6.65272623e-02 7.10602524e-03 5.72723746e-01 -9.12714005e-01
-1.09348759e-01 -1.28518641e-01 -6.22997046e-01 8.21853161e-01
3.09970349e-01 -2.79605091e-01 -1.44359016e+00 -4.77116764e-01
-1.28529340e-01 -7.04457045e-01 1.03859437e+00 2.57326990e-01
1.48193848e+00 3.78250293e-02 -2.21764565e-01 6.25997961e-01
1.31851888e+00 1.86966453e-02 2.68797755e-01 6.50431290e-02
8.47825825e-01 7.18180716e-01 1.12904429e+00 3.88769090e-01
3.81762087e-01 5.00464916e-01 2.16436297e-01 -3.33180308e-01
-7.14000106e-01 -2.72630453e-01 2.18211770e-01 8.10850799e-01
3.76327485e-01 -3.92567903e-01 -8.70005429e-01 4.42486465e-01
-1.86167312e+00 -6.67958319e-01 -4.93819922e-01 2.12855959e+00
1.22558248e+00 2.13930801e-01 -1.87145509e-02 4.44994479e-01
8.22383225e-01 -2.85733547e-02 -7.69348979e-01 3.79057080e-01
-4.47050363e-01 -5.87351471e-02 7.32947588e-01 2.58821875e-01
-1.40024221e+00 8.83538723e-01 5.12060881e+00 1.66392553e+00
-9.25796568e-01 6.23337388e-01 1.18885207e+00 -5.99787571e-02
-4.26512212e-02 -2.28748918e-01 -9.29492354e-01 8.11560154e-01
8.13832343e-01 2.69927710e-01 2.09250465e-01 7.64044821e-01
2.58208603e-01 -2.45740250e-01 -8.86253357e-01 1.20899940e+00
1.08353198e-01 -7.37355113e-01 -2.62218684e-01 -6.51847199e-02
9.78599548e-01 1.69834226e-01 2.79835284e-01 5.59843719e-01
4.11654681e-01 -7.34795868e-01 8.72455895e-01 1.41819537e-01
1.22019935e+00 -6.39882207e-01 8.70335460e-01 6.27092421e-01
-8.88978779e-01 -2.50671417e-01 -2.34203398e-01 1.27613395e-01
-7.26607591e-02 1.32285690e+00 -6.04905725e-01 3.53586018e-01
6.84310675e-01 8.60963166e-01 -8.39604259e-01 9.84538198e-01
-4.93485302e-01 1.27623534e+00 -2.58652091e-01 3.39898378e-01
1.04855634e-01 2.46831011e-02 1.37579575e-01 1.32298279e+00
4.47229035e-02 4.91775423e-02 4.09025788e-01 4.97575223e-01
-2.84952044e-01 1.69303358e-01 -2.92344749e-01 1.89503551e-01
7.31695116e-01 1.26954138e+00 -7.34831512e-01 -4.75911647e-01
-2.10616127e-01 6.71010315e-01 3.95671070e-01 5.91568053e-01
-1.00858676e+00 -2.30101153e-01 3.20667058e-01 5.96214719e-02
-1.03611261e-01 2.25589082e-01 -4.90108162e-01 -1.18886399e+00
-1.47309065e-01 -8.61653149e-01 5.26679873e-01 -7.35896826e-01
-1.66757262e+00 5.76127648e-01 -6.70391619e-02 -1.42112839e+00
2.52765566e-01 -3.82630020e-01 -1.69771314e-01 6.73345506e-01
-1.72572148e+00 -1.31400180e+00 -1.93924353e-01 4.51924235e-01
7.86119521e-01 -2.54182875e-01 5.46177387e-01 7.23194003e-01
-7.85375655e-01 9.61397052e-01 5.13113499e-01 7.62167498e-02
1.39833200e+00 -1.29064405e+00 5.50431013e-02 7.18156755e-01
1.29999191e-01 1.89071715e-01 5.71702361e-01 -4.86352682e-01
-4.85677153e-01 -1.57324016e+00 7.80750215e-01 -3.78627241e-01
6.36520684e-01 -5.16651750e-01 -8.93519402e-01 4.47383881e-01
-2.38686755e-01 5.15134096e-01 7.55161524e-01 3.53867635e-02
-7.35387743e-01 -3.57098877e-01 -1.25379312e+00 2.91514397e-01
1.22597980e+00 -5.94217777e-01 -1.79406896e-01 7.11988151e-01
7.34090447e-01 -2.23901778e-01 -6.67513311e-01 6.23033524e-01
2.09874853e-01 -7.49422014e-01 8.57401669e-01 -3.11321676e-01
2.05386311e-01 -4.63984311e-01 -3.42567295e-01 -1.51329088e+00
-4.23938245e-01 -2.12879404e-01 1.52686983e-01 1.63663828e+00
5.63905299e-01 -6.75985396e-01 6.55090153e-01 2.59355783e-01
-1.98307529e-01 -7.86038578e-01 -7.56178737e-01 -9.21431780e-01
1.35799199e-01 -5.75716913e-01 2.99590111e-01 1.37252498e+00
-2.81070262e-01 4.02781963e-01 -4.15377915e-01 8.58043507e-02
1.13642812e+00 -6.37705252e-02 4.57448065e-01 -1.27416229e+00
-2.08634704e-01 -1.23435631e-01 7.75103569e-02 -9.35426950e-01
4.30418372e-01 -1.18861377e+00 3.78481090e-01 -1.08978057e+00
4.93699193e-01 -1.01007509e+00 -7.26766467e-01 7.96871185e-01
-4.98787910e-01 7.53176987e-01 1.77818984e-01 4.20181870e-01
-9.98132765e-01 5.25173187e-01 1.22376764e+00 -2.64837772e-01
9.13574100e-02 8.54708701e-02 -6.71061099e-01 5.33722997e-01
9.63170946e-01 -1.00790870e+00 -6.63615108e-01 -1.68509632e-01
-1.43506572e-01 -4.69564945e-01 1.18592940e-01 -9.71817255e-01
8.32616631e-03 -1.78210307e-02 1.23552166e-01 -4.28490192e-01
2.61912104e-02 -6.74492121e-01 -2.99024582e-02 2.73941845e-01
-7.10160136e-01 -5.78414202e-01 -6.72830045e-02 5.58937967e-01
-1.36800021e-01 -3.12634349e-01 1.05993271e+00 1.79809541e-01
-6.24540925e-01 3.47243220e-01 9.13982280e-03 5.00893831e-01
9.57019031e-01 2.24811539e-01 -6.13026559e-01 -1.92781746e-01
-6.63436532e-01 4.80610490e-01 3.54718804e-01 2.58289725e-01
1.73529968e-01 -1.36975753e+00 -9.45365250e-01 9.76252705e-02
4.85929221e-01 2.28235230e-01 4.77571011e-01 8.78822446e-01
-5.55581413e-02 1.21763490e-01 3.48367870e-01 -7.42852449e-01
-1.16377985e+00 5.82579374e-01 1.55447260e-01 -3.25689346e-01
-1.59774438e-01 1.06231928e+00 4.22908664e-01 -8.04010451e-01
4.58418995e-01 -7.56808044e-03 -8.68492573e-02 1.68091297e-01
5.55594504e-01 2.94961214e-01 1.79562137e-01 -6.85467422e-01
-2.63278455e-01 4.49617177e-01 -1.07143134e-01 1.96302012e-01
1.02683973e+00 -3.26723129e-01 3.64868864e-02 7.11362600e-01
1.44940078e+00 -2.30858222e-01 -1.37658656e+00 -6.98789597e-01
-3.07632275e-02 -3.23251009e-01 3.01357687e-01 -1.17791760e+00
-1.18281949e+00 8.16066682e-01 8.31209958e-01 -7.44872540e-02
1.13254988e+00 5.97721599e-02 6.65960848e-01 2.00029537e-01
4.50451136e-01 -1.25741017e+00 4.95646521e-02 2.44503334e-01
5.33292472e-01 -1.83796847e+00 -1.07974252e-02 -6.41487479e-01
-7.45461464e-01 4.03550684e-01 5.13337374e-01 2.39159063e-01
9.40793872e-01 1.99517637e-01 4.37697053e-01 5.66276535e-02
-7.03772366e-01 -3.25300246e-01 2.07125172e-01 6.09081089e-01
3.02167028e-01 2.32876867e-01 -1.89014658e-01 7.07998753e-01
2.27264196e-01 -2.54389774e-02 1.78966984e-01 7.23379731e-01
-3.88601452e-01 -1.04299092e+00 -4.62025404e-01 5.57591796e-01
-4.23303485e-01 -2.66145021e-01 2.79405005e-02 4.97050047e-01
4.51610237e-01 1.30211484e+00 -2.34858766e-01 -4.48925018e-01
2.57488787e-01 1.90741539e-01 7.21519664e-02 -7.87134707e-01
-2.30602786e-01 3.03011507e-01 4.99391742e-02 -3.08026105e-01
-7.10363150e-01 -6.21428609e-01 -1.33228958e+00 -9.34041850e-03
-5.35098314e-01 1.53645962e-01 5.66521049e-01 9.64707851e-01
2.90117711e-01 5.76539218e-01 9.04913068e-01 -5.44739664e-01
-7.08741724e-01 -1.15926838e+00 -9.08509254e-01 6.65967166e-01
2.53435016e-01 -8.54277909e-01 -7.29006886e-01 1.99344605e-01] | [9.42251205444336, 3.8769142627716064] |
23899834-e6e5-4717-9264-16302535abfe | tt-nf-tensor-train-neural-fields | 2209.15529 | null | https://arxiv.org/abs/2209.15529v1 | https://arxiv.org/pdf/2209.15529v1.pdf | TT-NF: Tensor Train Neural Fields | Learning neural fields has been an active topic in deep learning research, focusing, among other issues, on finding more compact and easy-to-fit representations. In this paper, we introduce a novel low-rank representation termed Tensor Train Neural Fields (TT-NF) for learning neural fields on dense regular grids and efficient methods for sampling from them. Our representation is a TT parameterization of the neural field, trained with backpropagation to minimize a non-convex objective. We analyze the effect of low-rank compression on the downstream task quality metrics in two settings. First, we demonstrate the efficiency of our method in a sandbox task of tensor denoising, which admits comparison with SVD-based schemes designed to minimize reconstruction error. Furthermore, we apply the proposed approach to Neural Radiance Fields, where the low-rank structure of the field corresponding to the best quality can be discovered only through learning. | ['Luc van Gool', 'Konrad Schindler', 'Christos Sakaridis', 'Mikhail Usvyatsov', 'Anton Obukhov'] | 2022-09-30 | null | null | null | null | ['low-rank-compression'] | ['computer-code'] | [ 2.29391083e-01 -5.51290512e-02 1.30003422e-01 -3.63041401e-01
-9.09310400e-01 -2.84000486e-01 2.98087209e-01 -2.44486444e-02
-3.47924471e-01 6.06042087e-01 4.39454913e-01 -3.77931111e-02
-8.37947607e-01 -7.10914552e-01 -1.09324014e+00 -9.18798089e-01
-4.01148468e-01 7.00068250e-02 -3.30377400e-01 -2.03746587e-01
1.36033073e-01 5.61274827e-01 -1.58339810e+00 4.42561626e-01
8.70494246e-01 1.30111969e+00 4.23080117e-01 5.43370008e-01
3.77893567e-01 1.09723222e+00 -2.15368867e-01 -3.61613810e-01
4.48334962e-01 -3.27664465e-02 -1.01037204e+00 1.13362923e-01
1.06245983e+00 -3.26224655e-01 -5.38603008e-01 1.14424789e+00
3.54816079e-01 4.47070241e-01 7.37029016e-01 -6.82721853e-01
-5.41179597e-01 6.41528249e-01 -3.07049125e-01 2.62566119e-01
-1.89374879e-01 -2.42271617e-01 1.38106465e+00 -1.01106453e+00
5.96402764e-01 1.17541695e+00 1.04077744e+00 1.94886521e-01
-1.60111392e+00 -2.68132448e-01 -1.16355971e-01 2.16904372e-01
-1.30655015e+00 -4.78231341e-01 7.48637557e-01 -7.67993987e-01
6.47646248e-01 2.33212560e-01 2.55526453e-01 7.71127164e-01
1.06053352e-01 6.90956593e-01 1.10871494e+00 -3.37275952e-01
2.74368882e-01 -2.26070061e-01 3.11637551e-01 8.61047864e-01
2.06909195e-01 9.25066769e-02 -6.07839167e-01 -3.48530769e-01
7.10132778e-01 8.88245646e-04 -6.11011446e-01 -8.36238489e-02
-1.17126715e+00 9.14195240e-01 6.82134867e-01 4.67756301e-01
-7.82641590e-01 6.37279630e-01 2.85399854e-01 1.47775501e-01
8.67741227e-01 5.62564075e-01 -5.69876134e-01 1.88920662e-01
-1.14418876e+00 3.06244642e-01 4.22329545e-01 5.26156545e-01
1.08605170e+00 2.28932545e-01 -4.19407606e-01 1.03433299e+00
1.16446815e-01 2.79246449e-01 1.06802888e-01 -1.23161948e+00
3.86365622e-01 1.26019269e-01 -1.04313932e-01 -1.18949461e+00
-2.87694395e-01 -8.47618759e-01 -1.29491580e+00 6.31852522e-02
3.60023081e-01 -1.19890966e-01 -6.74106181e-01 1.63478851e+00
1.64891288e-01 4.96094078e-01 -4.19343896e-02 8.36861730e-01
5.61308444e-01 6.79369390e-01 -1.70577541e-01 9.68932882e-02
9.73983943e-01 -5.64350188e-01 -5.19981742e-01 1.46718845e-01
8.08210135e-01 -6.59716606e-01 9.78651404e-01 6.68618083e-01
-9.73364830e-01 -4.25517023e-01 -8.13106716e-01 -2.99787939e-01
6.67764172e-02 5.87229908e-01 9.41192865e-01 3.68843794e-01
-1.29127812e+00 1.40199411e+00 -7.48424113e-01 3.15569974e-02
6.27316177e-01 2.56483585e-01 -3.11999589e-01 -3.47033501e-01
-9.78373647e-01 5.30589104e-01 1.58376083e-01 3.52165639e-01
-1.02952576e+00 -9.81726766e-01 -6.02591634e-01 2.04167902e-01
1.03626058e-01 -7.36458242e-01 1.02219212e+00 -9.50141132e-01
-1.21651256e+00 4.70666796e-01 -7.11945966e-02 -5.01595557e-01
6.46608397e-02 -4.07812148e-01 3.69248167e-02 3.21077704e-01
2.13148832e-01 4.80662018e-01 1.12770879e+00 -1.31228411e+00
-4.86662596e-01 -3.96355957e-01 1.55902594e-01 -1.28036365e-01
-6.54256403e-01 -2.81964123e-01 -8.51403475e-02 -1.02679014e+00
2.05364078e-01 -7.54929602e-01 -5.75611711e-01 1.22556418e-01
-3.35757732e-01 -2.23282471e-01 2.89365053e-01 -9.11923528e-01
1.14362061e+00 -2.25842929e+00 4.38512772e-01 4.53577936e-01
5.26313305e-01 -1.19038783e-01 -2.76567429e-01 1.69899479e-01
-1.01674676e-01 -5.01999334e-02 -2.96055257e-01 -4.39134628e-01
-1.80268988e-01 3.96726876e-01 -5.38241744e-01 7.12430060e-01
3.20344388e-01 4.63259131e-01 -9.37519848e-01 -1.98811799e-01
-2.53633678e-01 7.48160660e-01 -8.18505168e-01 5.67230321e-02
-1.92837581e-01 4.18441683e-01 -5.35609722e-01 4.20434415e-01
7.80528605e-01 -5.47866762e-01 7.64502538e-03 -7.80116796e-01
-1.46698460e-01 2.64914781e-01 -1.20938993e+00 1.56394148e+00
-6.77032232e-01 6.85611963e-01 3.62847745e-01 -1.18048501e+00
6.80568635e-01 1.55529216e-01 7.55423844e-01 -4.54935610e-01
5.31711355e-02 3.10906678e-01 -3.61069828e-01 -4.20448035e-01
5.51734328e-01 -1.28126428e-01 4.36308086e-01 2.13585109e-01
2.29121014e-01 2.35326856e-01 3.64925802e-01 2.18766555e-01
1.25481117e+00 -9.18519944e-02 -1.99761331e-01 -5.59438586e-01
3.45194846e-01 -2.40680829e-01 3.21352333e-01 8.28304350e-01
4.61261481e-01 5.84481001e-01 3.85833323e-01 -3.93830448e-01
-8.74206007e-01 -6.56053901e-01 -4.79467779e-01 1.31060708e+00
-3.77252191e-01 -5.02028167e-01 -6.25201344e-01 -4.43436295e-01
1.98970214e-02 4.12175953e-01 -6.54504120e-01 8.98549631e-02
-6.16617858e-01 -7.93256283e-01 5.46929479e-01 3.14852297e-01
3.53396773e-01 -6.65967047e-01 -3.15161407e-01 7.87391588e-02
-3.93107265e-01 -1.03054285e+00 -3.87697846e-01 5.27073324e-01
-1.27967477e+00 -7.27297544e-01 -9.49632764e-01 -6.27514541e-01
7.75509059e-01 3.78340393e-01 1.25107050e+00 1.08207069e-01
1.02690518e-01 3.06297123e-01 -1.71364069e-01 1.22890219e-01
-5.10693453e-02 1.56788841e-01 -3.77194732e-02 4.82710421e-01
-3.28028530e-01 -6.55536890e-01 -4.80545163e-01 1.70223042e-02
-1.17466164e+00 -8.90712067e-02 5.30717790e-01 9.34981108e-01
8.54872644e-01 2.87267327e-01 1.66978821e-01 -8.68801057e-01
6.23521864e-01 -5.95488489e-01 -6.68769300e-01 1.39808998e-01
-3.70783061e-01 6.16429448e-01 8.05624247e-01 -3.32682520e-01
-8.47988069e-01 2.37772837e-01 -2.09143639e-01 -6.91620111e-01
1.06452122e-01 8.78638208e-01 3.06232542e-01 -5.82887828e-01
9.34859574e-01 6.85988814e-02 -2.93666244e-01 -8.03183675e-01
4.36701953e-01 1.95804760e-01 4.79900539e-01 -9.88605320e-01
7.95403957e-01 5.96432924e-01 5.12303472e-01 -9.62936640e-01
-1.42041266e+00 -5.32440186e-01 -4.82630074e-01 -2.58540481e-01
6.40752912e-01 -9.63612437e-01 -6.03989720e-01 2.66857207e-01
-1.29328108e+00 -3.26600850e-01 -4.39752460e-01 5.66217065e-01
-5.19123018e-01 3.35854948e-01 -6.59525573e-01 -5.85422873e-01
-3.36369812e-01 -1.07096493e+00 1.25946939e+00 -3.43688428e-01
4.52801675e-01 -1.05038452e+00 2.17343181e-01 2.41004810e-01
3.21256816e-01 6.56341389e-02 9.19665635e-01 -4.37958315e-02
-8.20452631e-01 1.65360048e-01 -3.67553204e-01 8.20723116e-01
-2.67138600e-01 -3.78156155e-01 -1.20341384e+00 -3.58637482e-01
8.64886418e-02 -4.51709747e-01 1.37626898e+00 7.84298778e-01
1.64365816e+00 -6.18659556e-01 1.46472082e-02 1.27000630e+00
1.59573376e+00 -5.01845241e-01 5.83150387e-01 1.75126702e-01
1.03310561e+00 7.33778059e-01 1.89070046e-01 5.19537508e-01
2.55850613e-01 4.90680397e-01 3.96206170e-01 -1.69549689e-01
-1.52412549e-01 1.16654865e-01 2.67067641e-01 8.56748819e-01
-4.92505431e-01 6.74198419e-02 -7.71142483e-01 4.96139854e-01
-1.74557281e+00 -9.32630599e-01 -2.30908051e-01 2.16125870e+00
8.26171875e-01 -2.33383179e-01 -1.24519929e-01 3.06011140e-01
4.16159898e-01 1.91039070e-01 -1.84950128e-01 9.24214125e-02
-2.56287038e-01 5.35470486e-01 7.38843203e-01 4.99587238e-01
-1.27178001e+00 5.92125356e-01 6.33782625e+00 9.27566350e-01
-1.36146200e+00 2.22172603e-01 7.42234111e-01 1.42079085e-01
-4.88340795e-01 -1.23512149e-01 -7.02158809e-01 1.08441703e-01
7.28788078e-01 2.38422930e-01 7.25558460e-01 7.73433268e-01
3.10642958e-01 2.98100442e-01 -1.13967299e+00 9.74085331e-01
-1.50459319e-01 -1.75076282e+00 1.30553380e-01 1.45831257e-01
1.01296270e+00 2.97969937e-01 2.44666561e-01 7.18072951e-02
1.75559431e-01 -1.02450550e+00 7.29356050e-01 6.85870409e-01
8.99471700e-01 -6.48409665e-01 5.32578588e-01 9.94421840e-02
-9.87397909e-01 -1.83405980e-01 -6.44678235e-01 1.12653881e-01
-7.87333399e-02 1.27968442e+00 -5.72252333e-01 3.98807943e-01
9.93304253e-01 1.01021481e+00 -5.41640043e-01 1.02210832e+00
2.76751697e-01 9.63050246e-01 -3.31439793e-01 2.27529138e-01
4.17405486e-01 -3.75658840e-01 4.94862378e-01 1.24327481e+00
4.75779980e-01 7.24835843e-02 1.10921033e-01 7.46392250e-01
-3.48792285e-01 1.14819281e-01 -5.27008057e-01 4.78231944e-02
-2.32467398e-01 1.27222550e+00 -4.97818321e-01 -7.20087290e-02
-1.83365822e-01 6.30845666e-01 5.42308152e-01 4.81477648e-01
-2.52886653e-01 -2.46593684e-01 6.51064873e-01 3.46108496e-01
6.35793984e-01 -3.27226251e-01 -3.77839833e-01 -1.29421663e+00
7.68734291e-02 -8.72593701e-01 2.34631330e-01 -6.14849031e-01
-1.55346239e+00 7.18082786e-01 6.41965419e-02 -1.11176991e+00
-1.01560233e-02 -7.72689283e-01 -2.80607730e-01 8.27655196e-01
-1.65209317e+00 -9.69419539e-01 -2.36342624e-01 6.53559208e-01
4.79359664e-02 -3.31350528e-02 4.42744732e-01 7.46810377e-01
-3.03830177e-01 4.41952825e-01 6.45960212e-01 -1.09068342e-02
2.38306195e-01 -1.26294255e+00 6.98992014e-02 8.35454822e-01
1.78611845e-01 5.87092280e-01 5.24594307e-01 -4.23183143e-01
-1.44698060e+00 -1.47377443e+00 8.21103215e-01 3.29707302e-02
7.27398992e-01 -2.53277004e-01 -1.13894320e+00 4.79982287e-01
-1.78905353e-01 3.08953345e-01 3.84080976e-01 3.82892877e-01
-4.90602463e-01 -4.51881409e-01 -9.10796106e-01 3.53204697e-01
8.99730444e-01 -5.96805155e-01 -1.06371902e-02 8.93937528e-01
5.91998100e-01 -1.58099264e-01 -1.04857326e+00 2.89162010e-01
2.22303480e-01 -7.73780286e-01 1.19403768e+00 -6.00728691e-01
7.53890157e-01 -1.78353623e-01 -5.22549212e-01 -1.49968398e+00
-6.83125794e-01 -7.34600484e-01 -2.20821440e-01 6.87769175e-01
1.80682480e-01 -2.52650887e-01 8.12553942e-01 -8.05212855e-02
-2.91625142e-01 -9.14230704e-01 -9.00469959e-01 -5.81143498e-01
2.50233486e-02 -6.23140693e-01 2.15031773e-01 8.98495078e-01
-7.00213373e-01 4.51203585e-01 -6.57270133e-01 3.57841581e-01
1.04918289e+00 -7.21494108e-02 2.83859938e-01 -1.34392416e+00
-6.44502401e-01 -3.06450218e-01 -2.75533348e-01 -1.45680881e+00
3.20203871e-01 -1.27453041e+00 2.69939840e-01 -1.24649251e+00
5.72214909e-02 -6.68220937e-01 -2.63108671e-01 3.46545160e-01
9.48353261e-02 2.79592723e-01 1.57109797e-01 2.73148954e-01
-4.53943372e-01 6.88413858e-01 1.49676645e+00 -2.89602488e-01
1.02129810e-01 3.41595449e-02 -6.18422985e-01 6.27431035e-01
4.27456021e-01 -6.32419646e-01 -2.66635865e-01 -8.54000628e-01
6.06394053e-01 1.29357874e-01 5.43036580e-01 -1.03293216e+00
1.30349219e-01 -1.41414642e-01 2.48782218e-01 -3.95290285e-01
3.25770944e-01 -7.63958752e-01 -6.19731322e-02 2.11766094e-01
-6.18940294e-01 -7.60246664e-02 -1.41788244e-01 5.36662400e-01
-2.00662911e-01 -6.25049829e-01 8.04489136e-01 -8.67838331e-04
-4.34745520e-01 5.89541674e-01 -1.33980155e-01 3.74991484e-02
1.75250590e-01 1.96162716e-01 -1.12176247e-01 -3.79157484e-01
-6.33567214e-01 -2.01915592e-01 -1.63463205e-01 -1.65174648e-01
6.93783522e-01 -1.30656779e+00 -9.76104915e-01 -1.55305965e-02
-2.20789425e-02 1.24452949e-01 2.14410067e-01 8.08226347e-01
-6.49441957e-01 2.28800029e-01 7.52604455e-02 -6.45132303e-01
-8.54311287e-01 2.03844413e-01 4.86179680e-01 -4.63316858e-01
-6.64834082e-01 9.97422278e-01 2.55097657e-01 -3.46467644e-01
4.29127365e-01 -5.51923871e-01 -2.48056576e-01 -6.34251386e-02
4.57699448e-01 6.09948933e-01 4.78558809e-01 -6.92790985e-01
-8.22904557e-02 5.22955298e-01 1.87966943e-01 4.44223024e-02
1.76645613e+00 2.10473806e-01 -3.85249555e-01 1.81270093e-01
1.56309438e+00 -1.71064377e-01 -1.19891131e+00 -4.49720502e-01
4.13343795e-02 -5.16915619e-01 9.23012137e-01 -3.49454045e-01
-1.47352934e+00 8.06836426e-01 6.99094296e-01 2.15618670e-01
1.29905486e+00 -6.33499920e-02 8.60097170e-01 9.07131433e-01
3.26940790e-02 -8.31495643e-01 2.63184428e-01 7.76825190e-01
1.13406861e+00 -1.01012850e+00 6.79581538e-02 -2.65707672e-01
-2.08769858e-01 1.12255907e+00 1.62424579e-01 -3.51860315e-01
1.08202350e+00 1.69470936e-01 -1.63786098e-01 -4.71032977e-01
-7.43188381e-01 -2.77016878e-01 7.59376943e-01 4.10487354e-01
5.15355766e-01 2.64099631e-02 1.14370450e-01 1.19642034e-01
-1.28755033e-01 -3.89527678e-02 2.61395037e-01 4.93091404e-01
-3.58566821e-01 -7.51154780e-01 -5.01434863e-01 9.28557813e-01
-7.28630722e-01 -3.81037056e-01 1.27251953e-01 -2.69611087e-02
2.35779911e-01 7.34312117e-01 1.01468721e-02 -4.74586219e-01
2.52756506e-01 -4.82252419e-01 5.81384063e-01 -4.55361545e-01
-5.91929972e-01 1.15925841e-01 -3.83026972e-02 -5.41076839e-01
-5.24646997e-01 -6.95780694e-01 -7.06965387e-01 -2.89888859e-01
-2.60101825e-01 3.22807021e-02 6.03090048e-01 9.06410456e-01
2.40914360e-01 4.51892138e-01 8.29126120e-01 -1.12343383e+00
-8.60164285e-01 -8.44393432e-01 -7.21551180e-01 4.88400280e-01
7.49310136e-01 -6.24544799e-01 -4.87301171e-01 1.95349991e-01] | [11.463295936584473, -2.1143999099731445] |
8e034986-4092-4897-a5d6-8a2a9f3db048 | relwalk-a-latent-variable-model-approach-to-2 | null | null | https://aclanthology.org/2021.eacl-main.133 | https://aclanthology.org/2021.eacl-main.133.pdf | RelWalk - A Latent Variable Model Approach to Knowledge Graph Embedding | Embedding entities and relations of a knowledge graph in a low-dimensional space has shown impressive performance in predicting missing links between entities. Although progresses have been achieved, existing methods are heuristically motivated and theoretical understanding of such embeddings is comparatively underdeveloped. This paper extends the random walk model of word embeddings to Knowledge Graph Embeddings (KGEs) to derive a scoring function that evaluates the strength of a relation R between two entities h (head) and t (tail). Moreover, we show that marginal loss minimisation, a popular objective used in much prior work in KGE, follows naturally from the log-likelihood ratio maximisation under the probabilities estimated from the KGEs according to our theoretical relationship. We propose a learning objective motivated by the theoretical analysis to learn KGEs from a given knowledge graph.Using the derived objective, accurate KGEs are learnt from FB15K237 and WN18RR benchmark datasets, providing empirical evidence in support of the theory. | ['Ken-ichi Kawarabayashi', 'Yuichi Yoshida', 'Huda Hakami', 'Danushka Bollegala'] | 2021-04-01 | null | null | null | eacl-2021-2 | ['knowledge-graph-embeddings', 'knowledge-graph-embeddings'] | ['graphs', 'methodology'] | [-1.10290535e-01 8.37892830e-01 -4.83672112e-01 -3.57740581e-01
-1.99252859e-01 -4.19159949e-01 6.09501243e-01 4.59161162e-01
-4.33886588e-01 7.13733673e-01 2.20856965e-01 -1.89938396e-01
-7.99547911e-01 -1.20477426e+00 -7.90492892e-01 -5.29631495e-01
-7.52828360e-01 5.05518079e-01 2.74319828e-01 -1.97959063e-03
-1.70190185e-01 3.74446481e-01 -1.05178797e+00 -4.05562103e-01
7.72405267e-01 7.65336275e-01 -1.32671073e-01 5.67237318e-01
-2.24682186e-02 6.99938536e-01 -1.81402057e-01 -1.14272892e+00
6.69050356e-03 -2.15072393e-01 -1.10447538e+00 -1.73261106e-01
2.66122371e-01 1.63128227e-01 -7.15163767e-01 9.80856955e-01
3.84443700e-01 2.30767205e-01 1.02692473e+00 -1.51505923e+00
-1.13088787e+00 6.00792587e-01 -3.43938887e-01 1.65110886e-01
2.52774417e-01 -5.30903399e-01 2.00569272e+00 -8.55646074e-01
9.38825071e-01 9.15066242e-01 7.47193336e-01 1.01418763e-01
-1.26939940e+00 -2.14268386e-01 -7.20510036e-02 5.42905807e-01
-1.71117687e+00 4.82430644e-02 4.57228273e-01 -3.63813728e-01
1.20901251e+00 -6.71910569e-02 6.25298440e-01 7.61651933e-01
1.03242053e-02 6.98255837e-01 6.00693226e-01 -6.08587444e-01
-2.94609666e-02 3.48212719e-01 3.95359814e-01 1.09241855e+00
8.62794101e-01 -8.44867453e-02 -8.39483857e-01 -6.75166994e-02
5.42337954e-01 -3.48878652e-01 -3.24798584e-01 -9.32615280e-01
-8.31163466e-01 9.28963542e-01 7.90700316e-01 2.47507244e-01
-3.52225870e-01 1.48831725e-01 2.18558803e-01 3.76953751e-01
3.61276299e-01 3.67822915e-01 -5.70715129e-01 1.29483789e-01
-3.89870286e-01 1.56086981e-01 1.12122583e+00 9.29992199e-01
8.13093901e-01 -3.00969720e-01 1.06372170e-01 6.98401153e-01
5.48753858e-01 2.24070415e-01 7.69310892e-02 -3.28379333e-01
3.40295702e-01 8.61605525e-01 6.36477992e-02 -1.47505903e+00
-2.98557550e-01 -5.26813745e-01 -6.67623103e-01 -1.91661060e-01
3.89289975e-01 -9.82204080e-02 -4.16419983e-01 1.78155482e+00
3.98141444e-01 3.73193443e-01 1.87277958e-01 5.92888892e-01
6.15484238e-01 3.60143751e-01 -2.68040933e-02 7.92302415e-02
1.06638908e+00 -6.07458532e-01 -7.26906240e-01 5.07884398e-02
8.38492930e-01 -1.45793080e-01 6.22273505e-01 1.43347338e-01
-6.64314210e-01 -7.09285513e-02 -1.02604127e+00 -1.06050316e-02
-7.96291351e-01 -1.43222988e-01 9.18447852e-01 7.37174213e-01
-1.05722272e+00 6.06186509e-01 -6.73503816e-01 -5.21869719e-01
4.93995637e-01 3.54471475e-01 -7.15318263e-01 -2.09905028e-01
-1.60495198e+00 1.22929454e+00 6.22707188e-01 1.03310697e-01
-4.09581631e-01 -6.44614756e-01 -1.04241741e+00 2.23227188e-01
5.68944275e-01 -7.79994249e-01 4.88879591e-01 -2.64354676e-01
-1.09960735e+00 8.49172652e-01 1.43487617e-01 -7.78999686e-01
1.76391706e-01 -1.92515597e-01 -6.47380531e-01 2.58664250e-01
-9.06428099e-02 3.89092803e-01 6.11364603e-01 -1.03369319e+00
-4.50628251e-01 -2.72863477e-01 1.74709573e-01 3.57979238e-02
-7.57338941e-01 -4.04593498e-01 -2.92981744e-01 -4.19335842e-01
-2.02808946e-01 -6.37148976e-01 1.46510780e-01 -7.24327238e-03
-6.13321543e-01 -6.05297327e-01 5.60419023e-01 -6.88353479e-01
1.47247720e+00 -1.86100614e+00 4.50567693e-01 6.49889469e-01
4.95018154e-01 4.14023250e-01 -6.86975643e-02 6.74298704e-01
-1.86445683e-01 1.38701111e-01 3.08158481e-03 5.63729554e-02
4.32172716e-01 3.41909766e-01 -2.41718858e-01 4.91897196e-01
2.48497695e-01 1.15251946e+00 -1.08954477e+00 -3.23144108e-01
8.31983984e-02 5.31196356e-01 -5.22065818e-01 9.44201201e-02
-4.05319110e-02 -6.31148458e-01 -3.91606450e-01 3.67388725e-01
3.77437145e-01 -5.52266836e-01 6.07342005e-01 -3.81810248e-01
3.48219573e-01 1.09588906e-01 -1.15127861e+00 1.33804071e+00
-1.92849770e-01 6.34867787e-01 -2.95889854e-01 -1.28400505e+00
8.86912584e-01 2.27831870e-01 3.09042394e-01 -3.03261817e-01
4.69344966e-02 -8.28538388e-02 -1.33732796e-01 -4.70525265e-01
4.62630659e-01 -9.96493101e-02 -5.78891374e-02 1.66377977e-01
6.06458426e-01 2.65112519e-01 1.89951152e-01 7.35092223e-01
1.51024902e+00 1.17703937e-01 5.72079182e-01 -5.05345240e-02
2.50941724e-01 -2.56654799e-01 3.20502609e-01 6.63604736e-01
-7.57681131e-02 4.29538265e-02 7.97330379e-01 -8.08862150e-02
-8.93658996e-01 -1.29060662e+00 -2.53545702e-01 7.87228942e-01
8.83478373e-02 -8.38174641e-01 -4.25255448e-01 -1.07282293e+00
4.44991231e-01 6.65704310e-01 -8.55196238e-01 -3.51348728e-01
-3.79476696e-02 -7.22329557e-01 5.39071381e-01 4.38034147e-01
1.51652396e-01 -6.02414608e-01 -4.63335216e-02 4.65288945e-02
1.36792421e-01 -1.17521989e+00 -2.33274713e-01 1.73908174e-01
-5.79922318e-01 -1.53978467e+00 -5.07665277e-01 -7.30771661e-01
5.92948794e-01 -3.99242416e-02 1.11346185e+00 -4.10417207e-02
-5.43016016e-01 8.43424082e-01 -4.89771664e-01 -4.31814007e-02
1.27704993e-01 5.77352270e-02 2.82660931e-01 -5.67538887e-02
6.72134817e-01 -6.58348024e-01 -2.95552582e-01 3.27159494e-01
-9.61592257e-01 -4.05802131e-01 7.52839088e-01 8.27464819e-01
4.53817308e-01 4.71881777e-01 9.01874423e-01 -9.64005530e-01
6.44785583e-01 -8.82613242e-01 -4.70250905e-01 5.39012849e-01
-1.12731552e+00 4.33386058e-01 2.97516584e-01 -1.19491912e-01
-7.46144354e-01 -2.86236942e-01 1.94376022e-01 -1.70740873e-01
1.68001577e-01 8.98635805e-01 -1.18092924e-01 -1.43662497e-01
4.34588879e-01 -1.05627082e-01 -3.40692788e-01 -2.77253956e-01
9.91787791e-01 4.41654116e-01 3.62400442e-01 -5.81186831e-01
1.06775689e+00 2.99767286e-01 2.97528386e-01 -9.22661960e-01
-1.12913764e+00 -5.49646497e-01 -7.30347216e-01 7.11401403e-02
7.34899521e-01 -7.53604531e-01 -8.81668866e-01 -1.78247645e-01
-8.31849098e-01 3.90791334e-02 -3.11930686e-01 7.41190076e-01
-3.75534385e-01 4.36776996e-01 -3.26279789e-01 -7.99096763e-01
-2.21336838e-02 -3.17284822e-01 5.65339386e-01 1.09108031e-01
-2.85687327e-01 -1.49813926e+00 3.99401963e-01 2.14064881e-01
6.81371540e-02 2.81278104e-01 1.30865610e+00 -8.63879263e-01
-5.96313536e-01 -5.53959668e-01 -4.88338321e-01 3.98255646e-01
-1.70787319e-03 -7.19040632e-02 -6.33272946e-01 -6.37324154e-02
-7.69547880e-01 -3.48636657e-01 9.27620530e-01 1.05099455e-01
5.76052666e-01 -2.37960711e-01 -6.14038289e-01 2.87277609e-01
1.68341494e+00 -5.11502147e-01 6.54525459e-01 3.26261133e-01
5.96582890e-01 6.00070775e-01 4.26904559e-01 2.65512586e-01
5.87814271e-01 6.49679005e-01 4.71368223e-01 3.44255507e-01
-5.48605956e-02 -6.33576810e-01 4.06614125e-01 8.87621522e-01
-1.72908127e-01 -5.34676015e-01 -8.11704338e-01 9.49178874e-01
-1.87523651e+00 -8.33716094e-01 -2.49046952e-01 2.19205928e+00
9.39709365e-01 2.40578130e-01 9.21766087e-02 -5.08573884e-03
6.31617427e-01 1.70129403e-01 -2.05530971e-01 -1.93935737e-01
-2.04997733e-01 3.62605989e-01 6.49736166e-01 5.89424372e-01
-7.70286202e-01 8.54166269e-01 5.72562838e+00 8.57405365e-01
-2.39682406e-01 -1.16900159e-02 -1.33687958e-01 2.44612381e-01
-4.24130052e-01 9.40511823e-02 -8.96663189e-01 1.51544094e-01
9.20415401e-01 -4.76832271e-01 1.52769014e-01 6.53806508e-01
-3.19234163e-01 6.00368017e-03 -1.18959737e+00 8.08251083e-01
1.44795686e-01 -1.14206982e+00 -3.69467214e-02 2.70772964e-01
8.09867203e-01 -3.96235771e-02 7.60798007e-02 5.31511307e-01
7.57778764e-01 -1.05509591e+00 -2.21686345e-03 8.13558638e-01
5.06351888e-01 -8.49581122e-01 8.84828031e-01 1.16837554e-01
-1.20049596e+00 1.73366889e-01 -5.43638766e-01 1.12665623e-01
1.41774684e-01 9.77161050e-01 -1.23360002e+00 1.04858398e+00
3.33190948e-01 8.41156185e-01 -5.82591534e-01 9.54111993e-01
-8.27797353e-01 7.35058308e-01 -2.28771880e-01 -5.97718284e-02
1.65519807e-02 -1.82727024e-01 6.82876348e-01 1.13401830e+00
3.59481224e-03 -2.18935832e-01 -2.52159894e-01 7.66327679e-01
-4.94822443e-01 2.74833530e-01 -7.57531106e-01 -6.00209594e-01
3.50769669e-01 1.31168878e+00 -5.09432316e-01 9.71644670e-02
-5.73083580e-01 9.70790744e-01 9.30752695e-01 3.61447901e-01
-6.38888359e-01 -7.95833349e-01 7.01556742e-01 5.31303734e-02
7.70976782e-01 -2.09128588e-01 3.34879637e-01 -9.67858851e-01
1.62323952e-01 -2.39712119e-01 6.64960325e-01 -5.64710796e-01
-1.68913972e+00 2.24503949e-01 6.49642125e-02 -6.69292450e-01
-5.20485677e-02 -7.87137747e-01 -3.36391866e-01 7.40894079e-01
-1.78534997e+00 -8.80790293e-01 9.31562111e-02 2.61596054e-01
-3.05368334e-01 -9.03072655e-02 1.12436152e+00 3.06397080e-01
-5.89777648e-01 7.16503263e-01 2.66827643e-01 4.25038666e-01
5.02634883e-01 -1.58891046e+00 2.37001210e-01 5.48121929e-01
4.93485451e-01 7.73626328e-01 5.99675179e-01 -8.29202294e-01
-1.50783491e+00 -1.09590781e+00 1.22785258e+00 -6.53551698e-01
1.20634580e+00 -2.99546957e-01 -8.89521778e-01 8.34793627e-01
5.11536859e-02 3.44805479e-01 9.14408147e-01 7.52481580e-01
-6.85773611e-01 6.64070621e-02 -1.00401568e+00 3.62040073e-01
1.13273537e+00 -6.55290663e-01 -8.08117151e-01 2.88042724e-01
5.46106279e-01 7.38937035e-02 -1.37085700e+00 3.01052064e-01
5.79856455e-01 -5.53579628e-01 1.15363717e+00 -1.05961931e+00
2.61091441e-01 -2.63556093e-01 -1.91884592e-01 -1.54407370e+00
-2.21288785e-01 -4.09411132e-01 -7.81921387e-01 1.17077112e+00
4.89626169e-01 -5.70255816e-01 8.99784386e-01 3.39457691e-01
3.38332504e-01 -1.03714240e+00 -8.50916862e-01 -1.10308957e+00
-2.03137398e-01 -2.44787186e-01 2.05790102e-01 1.23317790e+00
3.39803576e-01 7.20645726e-01 -3.62241358e-01 4.98854280e-01
8.37699950e-01 -1.07885413e-01 6.18912399e-01 -1.55196965e+00
-3.18112314e-01 -2.61666566e-01 -1.09372568e+00 -9.58914220e-01
5.01691639e-01 -1.32915568e+00 -4.04451042e-01 -1.93571901e+00
1.85340136e-01 -3.22611272e-01 -4.43372905e-01 4.19010848e-01
-1.37477756e-01 4.73787077e-02 -1.92347676e-01 -3.44705522e-01
-9.60054398e-01 7.23420799e-01 8.48521709e-01 -6.62293434e-02
1.85268655e-01 -1.85543552e-01 -6.12016559e-01 7.21812427e-01
5.14887333e-01 -4.82718438e-01 -5.55868983e-01 -1.12400152e-01
8.95347655e-01 -1.92491442e-01 4.78795618e-01 -3.90386552e-01
3.40513825e-01 3.27809364e-01 9.18462649e-02 -2.91983634e-01
1.70427456e-01 -9.89114583e-01 -1.53387152e-03 7.02342987e-02
-3.41263890e-01 -4.62829113e-01 -2.49975488e-01 1.35302353e+00
-1.81202739e-01 -3.29140902e-01 1.68518990e-01 3.24119776e-01
-9.34748173e-01 3.41142327e-01 2.81738371e-01 4.02875960e-01
9.39561784e-01 -7.68203065e-02 -2.31328860e-01 -3.41776073e-01
-1.12165666e+00 2.68342853e-01 2.71888021e-02 3.23081493e-01
6.37170017e-01 -1.44396341e+00 -5.59921145e-01 -3.17221850e-01
4.10478979e-01 -2.57570118e-01 2.91683013e-03 8.24666739e-01
-8.34915861e-02 4.66382563e-01 2.40193978e-01 2.96643865e-03
-1.04504728e+00 4.65169400e-01 8.07804018e-02 -5.57052910e-01
-5.84877670e-01 1.08741581e+00 -2.58449495e-01 -4.22151327e-01
3.68367553e-01 -4.82555851e-03 -2.31531858e-01 2.98751831e-01
1.60098180e-01 6.30757391e-01 9.63396877e-02 -5.44555426e-01
-4.04553890e-01 2.48885915e-01 -2.75490284e-01 1.90084994e-01
1.64524913e+00 -5.68042994e-02 -1.45240594e-02 2.85390705e-01
1.45048821e+00 9.68266428e-02 -7.59241104e-01 -7.59903967e-01
4.51102018e-01 -5.10295093e-01 7.17466027e-02 -6.13462985e-01
-6.84488893e-01 5.21414161e-01 2.29228675e-01 5.19844353e-01
5.12343705e-01 3.36850256e-01 5.96565366e-01 7.03594923e-01
3.44238967e-01 -1.06107593e+00 -8.98299366e-02 3.93589795e-01
4.54353422e-01 -1.06209612e+00 2.15007961e-01 -5.96360207e-01
-4.32692617e-01 9.62386608e-01 1.79653153e-01 -1.53887257e-01
9.75618839e-01 -3.77037898e-02 -6.85518861e-01 -6.36763275e-01
-6.36385679e-01 -5.22418559e-01 5.17620802e-01 8.93444002e-01
2.40131781e-01 1.88394234e-01 -2.73110390e-01 6.57240450e-01
-1.32213369e-01 -1.02587104e-01 2.75974482e-01 6.19735420e-01
-5.08236110e-01 -1.16842890e+00 3.39952141e-01 5.89427233e-01
-3.53412479e-01 -2.33201548e-01 -4.17353779e-01 8.94067466e-01
-3.04067545e-02 6.49665356e-01 -2.28095144e-01 -3.35484177e-01
2.67565608e-01 2.94292897e-01 6.09322369e-01 -6.52145684e-01
1.89875275e-01 -7.18735933e-01 3.00845832e-01 -1.50538251e-01
-5.13362169e-01 -5.05296707e-01 -1.13047957e+00 -1.92131326e-01
-8.02694917e-01 3.67835492e-01 4.63898063e-01 9.42179322e-01
3.40919733e-01 5.98076761e-01 3.44281405e-01 1.00114895e-02
-4.15561348e-01 -6.71378672e-01 -1.15004814e+00 4.38663185e-01
-4.72187735e-02 -9.50100541e-01 -4.00242418e-01 -1.38134316e-01] | [8.748235702514648, 7.782365798950195] |
26ccb2e1-01e4-4698-ab99-345c21c3f784 | memory-efficient-tries-for-sequential-pattern | 2202.06834 | null | https://arxiv.org/abs/2202.06834v1 | https://arxiv.org/pdf/2202.06834v1.pdf | Memory Efficient Tries for Sequential Pattern Mining | The rapid and continuous growth of data has increased the need for scalable mining algorithms in unsupervised learning and knowledge discovery. In this paper, we focus on Sequential Pattern Mining (SPM), a fundamental topic in knowledge discovery that faces a well-known memory bottleneck. We examine generic dataset modeling techniques and show how they can be used to improve SPM algorithms in time and memory usage. In particular, we develop trie-based dataset models and associated mining algorithms that can represent as well as effectively mine orders of magnitude larger datasets compared to the state of the art. Numerical results on real-life large-size test instances show that our algorithms are also faster and more memory efficient in practice. | ['Andre A. Cire', 'Willem-Jan van Hoeve', 'Amin Hosseininasab'] | 2022-02-06 | null | null | null | null | ['sequential-pattern-mining'] | ['natural-language-processing'] | [ 3.45006824e-01 -1.31082803e-01 -5.41465938e-01 -2.15144888e-01
-8.13278556e-02 -3.29527020e-01 1.18477575e-01 5.03347397e-01
-2.99984425e-01 9.12605643e-01 -3.67907405e-01 -5.62091589e-01
-8.13273430e-01 -1.13199413e+00 -4.04669553e-01 -3.47718269e-01
-4.39595014e-01 9.85440135e-01 3.17142725e-01 1.88166380e-01
4.34085667e-01 5.95839798e-01 -2.06037498e+00 5.89331865e-01
8.44042659e-01 9.92450774e-01 1.25989616e-01 2.95575082e-01
-4.59393084e-01 6.77841485e-01 -5.37221551e-01 -3.04308776e-02
3.01864564e-01 -3.86644015e-03 -1.14975405e+00 2.82822281e-01
1.19140120e-02 3.80927831e-01 -1.59556568e-01 6.86094522e-01
4.86971438e-02 1.75761059e-01 2.21016437e-01 -1.22427201e+00
-4.57418039e-02 8.10712457e-01 -1.04198194e+00 5.55786788e-01
2.46332884e-01 -5.18314719e-01 6.90157235e-01 -9.03674603e-01
6.86526000e-01 8.77997279e-01 4.84769106e-01 3.08361828e-01
-1.16067493e+00 -9.32495177e-01 1.81609970e-02 6.19183838e-01
-1.62438369e+00 -2.05700725e-01 5.90524197e-01 -2.10628793e-01
1.18115377e+00 5.40763915e-01 6.59872890e-01 3.40749860e-01
-4.32367213e-02 8.32330525e-01 9.32231307e-01 -8.32099319e-01
1.87905028e-01 3.21397930e-01 7.87490726e-01 5.45243859e-01
8.11485529e-01 -2.01065447e-02 -8.98752153e-01 -3.53937238e-01
5.12971401e-01 1.27976194e-01 7.69860074e-02 -3.32606971e-01
-1.02316880e+00 9.60181773e-01 -3.42539310e-01 3.67437333e-01
-3.64375591e-01 -4.40705627e-01 2.63486981e-01 6.17346764e-01
2.91964650e-01 5.67949593e-01 -6.20504916e-01 -2.59615988e-01
-1.03507602e+00 3.05407226e-01 1.20211661e+00 1.09265602e+00
8.55357826e-01 -1.43279746e-01 2.77262509e-01 8.18608820e-01
-1.93945363e-01 9.07418802e-02 4.63785559e-01 -5.19482017e-01
3.50709200e-01 1.13609171e+00 9.88886654e-02 -9.90235686e-01
-7.36176968e-01 -5.22935271e-01 -7.64781117e-01 -2.05488563e-01
2.40400136e-01 1.28016472e-01 -6.12300396e-01 1.16846585e+00
4.26282495e-01 1.28072634e-01 4.80033830e-02 3.17629844e-01
2.10355863e-01 4.99400467e-01 -1.16926678e-01 -9.35788572e-01
1.27663529e+00 -7.03732491e-01 -6.17716312e-01 -1.89034134e-01
9.27463055e-01 -5.25374413e-01 6.10793948e-01 9.25898433e-01
-1.08632767e+00 -5.02583086e-01 -9.63280141e-01 4.29152459e-01
-4.22661841e-01 -2.24510163e-01 1.08188510e+00 6.92290008e-01
-3.79925817e-01 7.05602407e-01 -9.75054801e-01 -3.67410183e-01
7.43476391e-01 5.00244021e-01 -2.48650342e-01 -9.42527428e-02
-7.75539637e-01 6.06225312e-01 9.84523892e-01 -3.76070738e-01
-1.85893297e-01 -1.10904443e+00 -4.13437068e-01 -4.19650488e-02
7.99567044e-01 -3.33513379e-01 1.16194105e+00 -4.83613968e-01
-7.64966190e-01 8.91553402e-01 -5.77560604e-01 -8.74992847e-01
3.49233672e-02 -1.30291775e-01 -9.12328243e-01 -9.74888727e-02
-1.30414531e-01 7.09940717e-02 2.91006178e-01 -7.57185876e-01
-1.12528181e+00 -6.14590585e-01 -5.77976584e-01 -5.82671426e-02
-8.82500052e-01 9.23283175e-02 -2.51462609e-01 -4.67236102e-01
5.90961501e-02 -5.82383037e-01 -4.53893960e-01 -3.12135458e-01
-1.98973939e-01 -5.46312392e-01 1.13962972e+00 -1.62018210e-01
1.87303019e+00 -1.93479490e+00 -1.41150415e-01 8.35128307e-01
2.51474798e-01 1.84957549e-01 3.09998840e-01 6.11978650e-01
1.66192815e-01 4.02520411e-03 -2.30323687e-01 1.26031488e-01
-3.96943569e-01 7.88067937e-01 -3.94308835e-01 2.41081625e-01
-7.66914412e-02 5.33410728e-01 -7.73244381e-01 -4.88981247e-01
-1.40717172e-03 -1.87698901e-01 -4.14495468e-01 1.45261399e-02
-2.40433246e-01 8.81227776e-02 -3.34570289e-01 6.35658741e-01
5.10307014e-01 -6.04844153e-01 6.15809202e-01 2.27462947e-01
-4.22650784e-01 1.41365483e-01 -1.64350700e+00 1.22157371e+00
-8.66187960e-02 3.71573687e-01 -1.62171796e-01 -1.43186593e+00
1.12671697e+00 -1.01236977e-01 5.63948274e-01 -8.57774138e-01
-1.23793021e-01 3.22356522e-01 8.75256732e-02 -4.52063560e-01
4.88807946e-01 7.03773871e-02 1.12023763e-01 8.92384350e-01
-4.43758368e-01 5.16769648e-01 8.56113553e-01 -5.46322316e-02
1.21640253e+00 -7.69189179e-01 6.91220284e-01 -7.11651683e-01
5.70976079e-01 6.13158226e-01 7.42352784e-01 8.47921789e-01
5.48268080e-01 -1.67054147e-01 1.57298863e-01 -8.51059496e-01
-7.91792572e-01 -7.35922813e-01 -4.62858260e-01 1.02246284e+00
4.06701630e-03 -7.56370723e-01 -3.11088353e-01 -3.85975957e-01
4.67146814e-01 4.50153172e-01 -5.45603871e-01 2.51647651e-01
-6.52481198e-01 -1.09020746e+00 2.17680350e-01 7.81457484e-01
2.31967464e-01 -1.02082741e+00 -9.57325280e-01 5.58666348e-01
2.85603166e-01 -1.12694728e+00 2.31677398e-01 4.39021736e-01
-1.13133872e+00 -1.52374554e+00 1.51699170e-01 -7.85649478e-01
6.79661751e-01 3.94099325e-01 1.21807015e+00 2.07205996e-01
-8.40734005e-01 6.56835511e-02 -5.48244178e-01 -1.08119345e+00
-2.55251855e-01 2.94888794e-01 3.94517064e-01 -1.05253331e-01
1.06308210e+00 -6.96142614e-01 -1.62510917e-01 4.68272537e-01
-1.00989640e+00 1.05226964e-01 5.28831780e-01 7.53557503e-01
9.59740162e-01 8.78861368e-01 6.83100462e-01 -1.39829504e+00
5.88453352e-01 -8.09675217e-01 -8.87985468e-01 2.93352425e-01
-1.26356053e+00 3.38530131e-02 4.33457494e-01 -5.23022413e-01
-8.02951872e-01 5.45041561e-02 3.12245995e-01 -2.64835209e-01
-1.57184094e-01 9.37505007e-01 -3.69262323e-02 -5.04415296e-02
6.55291855e-01 4.41758871e-01 1.17651373e-01 -8.32534552e-01
-5.17174266e-02 8.07572722e-01 5.11340976e-01 -5.51877677e-01
7.55750120e-01 6.96065187e-01 3.46268088e-01 -1.07344210e+00
-8.62552941e-01 -8.35098803e-01 -6.83736205e-01 2.09479108e-01
1.47532970e-01 -5.05671799e-01 -7.44324863e-01 1.31680757e-01
-5.87383747e-01 -2.48614371e-01 -4.09135550e-01 1.54847249e-01
-3.33155274e-01 2.83377200e-01 -3.25205177e-01 -8.51449668e-01
-5.68180740e-01 -2.89144516e-01 3.04690063e-01 1.34948701e-01
-5.59020400e-01 -9.13486004e-01 3.24917436e-01 1.74739063e-01
1.79727703e-01 1.56039864e-01 1.01010084e+00 -9.89291012e-01
-3.09963137e-01 -5.77334017e-02 -1.75823346e-01 -1.18459657e-01
2.23786265e-01 -2.35795796e-01 -4.81552958e-01 -3.08408707e-01
-1.09935105e-01 -8.07484891e-03 6.46409631e-01 -8.83123502e-02
1.83625889e+00 -3.40391606e-01 -8.57125819e-01 4.19301629e-01
1.39237559e+00 4.22886044e-01 5.47453046e-01 5.18852532e-01
3.54517043e-01 7.84588397e-01 1.13951361e+00 9.67094898e-01
1.26996627e-02 4.86152023e-01 -1.64003074e-01 2.65851226e-02
3.49604338e-01 -1.39347434e-01 -3.21948797e-01 7.37122357e-01
-4.37115543e-02 8.96369666e-03 -1.21234679e+00 7.79056787e-01
-2.13317609e+00 -1.18441105e+00 -6.29221201e-01 2.17589307e+00
1.04878116e+00 3.40698063e-01 4.94064212e-01 1.03706610e+00
3.67523700e-01 -4.57340837e-01 -3.30574572e-01 -5.72114170e-01
-1.53452724e-01 7.21581340e-01 5.36585808e-01 8.55712686e-03
-9.51470673e-01 6.71771109e-01 7.14166737e+00 8.49540293e-01
-6.93034351e-01 -1.78465933e-01 3.53673220e-01 -3.60403121e-01
-1.38526410e-02 -2.84665287e-01 -1.13800812e+00 3.19434911e-01
1.28612077e+00 -6.87366426e-01 1.19071685e-01 1.05189431e+00
6.27870187e-02 -2.50556976e-01 -1.29229331e+00 1.25305867e+00
-2.01605350e-01 -1.86053658e+00 2.25433826e-01 3.88932198e-01
9.28236604e-01 -2.42647156e-01 -2.63106108e-01 -7.93272853e-02
7.79820010e-02 -1.02125573e+00 2.37190276e-01 3.93350482e-01
5.81947982e-01 -1.20331919e+00 5.48616827e-01 6.79763377e-01
-1.31525016e+00 -5.06661713e-01 -5.14821470e-01 -6.02963388e-01
-4.41105254e-02 1.01736784e+00 -7.61704803e-01 6.72802985e-01
9.67576981e-01 7.49991000e-01 -3.10243875e-01 1.23968768e+00
3.40415567e-01 7.47403026e-01 -7.39208877e-01 -1.43661395e-01
-8.50891918e-02 1.60013229e-01 2.04977021e-01 1.43043172e+00
1.50710076e-01 2.44782522e-01 3.86531383e-01 4.11055297e-01
5.15211858e-02 2.46512175e-01 -6.12359881e-01 -2.11662948e-01
9.28696990e-01 9.78696287e-01 -8.42322588e-01 -4.36173648e-01
-5.85093975e-01 3.22730571e-01 4.34882611e-01 -1.78429320e-01
-3.37985188e-01 -3.74714196e-01 7.37709463e-01 4.90954220e-01
4.65844750e-01 -3.11961383e-01 -6.62494659e-01 -7.12504268e-01
7.23461211e-02 -9.99487758e-01 1.07130992e+00 2.80815005e-01
-1.40181112e+00 2.47395903e-01 2.10836112e-01 -9.82022226e-01
-2.46045575e-01 -6.52076125e-01 -4.54946339e-01 4.82198298e-01
-1.44401586e+00 -5.26209116e-01 -2.34456077e-01 7.96456754e-01
4.09001827e-01 -2.13002369e-01 9.11735952e-01 3.44875783e-01
-7.11478889e-01 8.44068885e-01 2.67220646e-01 -1.08275630e-01
9.50280353e-02 -9.46085155e-01 1.64865151e-01 7.30411589e-01
4.89342958e-01 6.11821532e-01 5.87777138e-01 -6.53990030e-01
-1.75208712e+00 -1.08566236e+00 7.17155099e-01 -4.36094910e-01
7.64278710e-01 -1.89320281e-01 -1.23898363e+00 4.81966227e-01
-4.01932091e-01 -6.36918545e-02 1.32042408e+00 4.83115494e-01
-3.07770848e-01 -2.73191422e-01 -1.10167515e+00 8.51823241e-02
1.33008707e+00 -1.34250581e-01 -6.66798472e-01 1.52850553e-01
1.10952139e-01 -1.51671410e-01 -1.08920062e+00 6.55080080e-01
5.76259375e-01 -6.78793907e-01 6.54868960e-01 -7.76328921e-01
-4.80945222e-02 -8.39435160e-02 1.69485599e-01 -4.47197437e-01
-2.81270534e-01 -6.47874832e-01 -8.33884656e-01 8.28745902e-01
4.87633705e-01 -5.35938025e-01 1.32723522e+00 4.55852717e-01
3.51892889e-01 -1.16362023e+00 -7.85588205e-01 -1.17102003e+00
-1.70353934e-01 -6.89098537e-01 6.76816702e-01 1.16766691e+00
7.10145295e-01 1.70140222e-01 -2.17419922e-01 -8.16763565e-02
9.70221281e-01 7.75529861e-01 8.07139456e-01 -1.97678387e+00
-3.11087042e-01 -4.86654907e-01 -3.39793473e-01 -8.63077104e-01
-1.88005701e-01 -9.35216546e-01 -5.65861046e-01 -1.26881933e+00
4.10197586e-01 -7.40708709e-01 -4.52127635e-01 6.51552081e-01
1.38512060e-01 3.29484232e-02 -3.48019540e-01 4.08107966e-01
-6.74178064e-01 -1.41277224e-01 6.27102733e-01 8.39640647e-02
-4.57933813e-01 -2.29341574e-02 -6.88149691e-01 6.09893143e-01
1.05492139e+00 -6.31043553e-01 -4.16254371e-01 -1.11436300e-01
2.80600399e-01 -2.50910640e-01 -1.38927996e-01 -1.02413225e+00
6.24143779e-01 -4.53350365e-01 4.38865393e-01 -1.00822294e+00
-1.98429108e-01 -8.67805719e-01 4.26817477e-01 7.49125123e-01
-1.02586068e-01 6.58931211e-02 5.18343091e-01 6.97937012e-01
-2.00591400e-01 -1.71779573e-01 5.13892710e-01 -7.06429884e-04
-9.60186839e-01 3.56576353e-01 -3.03640544e-01 -2.09309399e-01
1.33313584e+00 -6.46768630e-01 -6.88666105e-02 4.10300434e-01
-6.07902944e-01 5.43379188e-01 8.01372975e-02 3.82750839e-01
8.21559906e-01 -1.03359234e+00 -6.67663932e-01 5.15584707e-01
3.50816607e-01 1.45823792e-01 6.50072545e-02 6.37989700e-01
-2.29152724e-01 5.84365726e-01 -1.17015339e-01 -4.54862505e-01
-1.66244292e+00 9.05565143e-01 -2.40950093e-01 -6.13570750e-01
-7.26257563e-01 6.17828786e-01 -2.72027880e-01 -6.03320310e-03
4.14151222e-01 7.84140676e-02 -1.24376781e-01 2.44082976e-02
1.27167678e+00 8.51581037e-01 4.09255065e-02 1.78822592e-01
-4.76714760e-01 2.99306840e-01 -5.19016445e-01 5.07009268e-01
1.44679058e+00 1.83604658e-01 -4.04271960e-01 5.24317980e-01
7.46248305e-01 -2.01785088e-01 -4.68590230e-01 -5.60462832e-01
8.48158240e-01 -5.15994191e-01 -1.89172044e-01 -5.68661749e-01
-8.79275620e-01 3.35056245e-01 3.13733578e-01 6.08968496e-01
1.22488713e+00 2.49250591e-01 7.68765807e-01 7.79795408e-01
7.58993804e-01 -1.38913047e+00 -1.86799258e-01 2.69861013e-01
5.02260506e-01 -9.02911961e-01 2.25894943e-01 -7.79704630e-01
-5.54554760e-02 1.25536931e+00 6.69449210e-01 4.56031188e-02
1.06904280e+00 8.59027505e-01 -4.52896893e-01 -4.08205450e-01
-1.16147292e+00 -1.46853939e-01 1.00626387e-01 5.40574908e-01
4.35466841e-02 1.04826249e-01 -6.28167272e-01 7.86496222e-01
-2.43359640e-01 3.38496000e-01 2.17768401e-01 1.42801094e+00
-7.65073538e-01 -1.47379470e+00 -3.71337503e-01 1.05378151e+00
-4.76887167e-01 1.92578137e-01 -4.05709773e-01 8.58100057e-01
3.28653120e-02 1.05145693e+00 4.47889715e-01 -5.50388396e-01
2.52731770e-01 2.22095102e-01 5.76354861e-01 -7.16029167e-01
-2.96157569e-01 -2.50059158e-01 2.94075608e-01 -5.29166996e-01
-4.35470819e-01 -6.58461988e-01 -1.53963506e+00 -6.62253499e-01
-3.13304424e-01 4.23275024e-01 2.65422523e-01 9.38555002e-01
5.55496812e-01 1.37755156e-01 5.38388193e-01 3.51173617e-02
-4.14665490e-01 -7.79526949e-01 -7.74083912e-01 2.37100482e-01
-2.86731690e-01 -6.89817131e-01 -6.78366125e-02 -1.76094800e-01] | [8.304780006408691, 6.278511047363281] |
6439de6a-27f5-4bf7-adac-663742200371 | sketch-guided-object-localization-in-natural | 2008.06551 | null | https://arxiv.org/abs/2008.06551v1 | https://arxiv.org/pdf/2008.06551v1.pdf | Sketch-Guided Object Localization in Natural Images | We introduce the novel problem of localizing all the instances of an object (seen or unseen during training) in a natural image via sketch query. We refer to this problem as sketch-guided object localization. This problem is distinctively different from the traditional sketch-based image retrieval task where the gallery set often contains images with only one object. The sketch-guided object localization proves to be more challenging when we consider the following: (i) the sketches used as queries are abstract representations with little information on the shape and salient attributes of the object, (ii) the sketches have significant variability as they are hand-drawn by a diverse set of untrained human subjects, and (iii) there exists a domain gap between sketch queries and target natural images as these are sampled from very different data distributions. To address the problem of sketch-guided object localization, we propose a novel cross-modal attention scheme that guides the region proposal network (RPN) to generate object proposals relevant to the sketch query. These object proposals are later scored against the query to obtain final localization. Our method is effective with as little as a single sketch query. Moreover, it also generalizes well to object categories not seen during training and is effective in localizing multiple object instances present in the image. Furthermore, we extend our framework to a multi-query setting using novel feature fusion and attention fusion strategies introduced in this paper. The localization performance is evaluated on publicly available object detection benchmarks, viz. MS-COCO and PASCAL-VOC, with sketch queries obtained from `Quick, Draw!'. The proposed method significantly outperforms related baselines on both single-query and multi-query localization tasks. | ['Anirban Chakraborty', 'Aditay Tripathi', 'Rajath R Dani', 'Anand Mishra'] | 2020-08-14 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/6490_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123510528.pdf | eccv-2020-8 | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 4.21815738e-02 -3.68892729e-01 -2.50974417e-01 -2.80093759e-01
-1.36295283e+00 -8.66101921e-01 9.52055275e-01 2.94961594e-02
-3.76490980e-01 3.44500214e-01 -7.72930011e-02 2.31736884e-01
1.61460899e-02 -5.26593745e-01 -9.48417246e-01 -6.78007960e-01
2.67961413e-01 6.72031164e-01 5.35541892e-01 -8.75302702e-02
4.97266412e-01 8.03788185e-01 -1.43476140e+00 3.48619848e-01
3.64779323e-01 1.17939806e+00 6.49020076e-01 4.99758631e-01
-1.95038468e-01 3.53662938e-01 -9.66435134e-01 -2.89483458e-01
2.17433065e-01 -1.69767529e-01 -7.92570174e-01 1.58214480e-01
1.15471041e+00 -3.57092917e-01 -3.08704525e-01 1.07117093e+00
4.65423405e-01 1.29093543e-01 8.16341102e-01 -1.41987431e+00
-1.06617212e+00 3.99048813e-02 -7.39002168e-01 1.78104967e-01
4.77910221e-01 3.81348245e-02 1.17970824e+00 -1.62842655e+00
8.18507314e-01 1.61669934e+00 1.59114376e-01 5.04561067e-01
-1.35537350e+00 -7.08192587e-01 4.82494056e-01 1.00271970e-01
-1.92188478e+00 -4.15659219e-01 9.38595057e-01 -2.97334939e-01
4.90054429e-01 2.58129328e-01 1.73520401e-01 1.28437614e+00
-1.82237998e-01 1.25817013e+00 6.91754460e-01 -3.18257153e-01
2.47140244e-01 3.41816604e-01 -1.11894086e-01 6.86743379e-01
3.99319157e-02 -6.97860345e-02 -5.25878966e-01 -4.58425522e-01
9.62803960e-01 4.01794165e-01 -3.33461940e-01 -7.17164040e-01
-1.50450575e+00 7.83994436e-01 6.68524265e-01 2.62648106e-01
-3.46083581e-01 3.89686793e-01 2.37223998e-01 1.47857726e-01
1.86058372e-01 4.02432173e-01 -2.27074653e-01 4.11452770e-01
-9.65038657e-01 5.05828500e-01 5.99163413e-01 1.39784622e+00
8.74828875e-01 -2.59092599e-01 -5.02292216e-01 1.00006890e+00
2.57446587e-01 7.52225935e-01 2.72313923e-01 -6.71342254e-01
6.34472549e-01 5.39301217e-01 5.18947005e-01 -1.13643825e+00
3.81223321e-01 -3.01938057e-01 -4.86046553e-01 1.62335575e-01
3.84900630e-01 4.77192879e-01 -1.02891839e+00 1.75417221e+00
1.88067660e-01 -4.55937646e-02 -1.34022743e-01 1.10658848e+00
1.10292661e+00 6.06461287e-01 1.70417324e-01 4.70043153e-01
1.52778780e+00 -1.16939807e+00 -4.91830438e-01 -3.96372020e-01
-4.56340313e-02 -8.26233983e-01 1.29467750e+00 8.06667358e-02
-9.66271579e-01 -7.96080291e-01 -8.96353126e-01 -1.77089497e-01
-8.15611184e-01 7.02545106e-01 3.76253098e-01 1.85048267e-01
-9.19047236e-01 9.06290561e-02 -3.72146428e-01 -5.90031683e-01
7.38229275e-01 2.55190253e-01 -5.80156267e-01 -4.89444405e-01
-7.74615526e-01 6.30428731e-01 2.75115311e-01 4.36302200e-02
-1.30003369e+00 -5.91699362e-01 -8.13184261e-01 1.57793105e-01
6.04528368e-01 -5.88372469e-01 1.09114575e+00 -9.76966619e-01
-9.30086613e-01 9.94881630e-01 -4.05156195e-01 2.20142603e-02
5.04341066e-01 -2.12049469e-01 -1.82332411e-01 3.53821605e-01
5.38209677e-01 1.17160225e+00 1.21374595e+00 -1.59545779e+00
-7.47767389e-01 -3.28232557e-01 7.96506405e-02 1.72882944e-01
1.95465803e-01 9.46471021e-02 -1.07407200e+00 -9.74072635e-01
1.16803020e-01 -7.56396890e-01 4.12941352e-02 5.85442007e-01
-4.19404864e-01 -6.26711726e-01 1.19993150e+00 -1.65847823e-01
7.74333656e-01 -2.21025181e+00 8.40719044e-02 3.05303335e-02
6.45923167e-02 2.60966837e-01 -5.04441738e-01 6.54601574e-01
8.51110145e-02 1.36219934e-01 -1.50912423e-02 -4.00632441e-01
1.07627623e-01 2.21132576e-01 -7.42187679e-01 3.52848351e-01
4.76109475e-01 1.37342453e+00 -1.03903186e+00 -5.61718047e-01
1.34325892e-01 3.52483213e-01 -1.60867929e-01 5.13622046e-01
-1.85673267e-01 1.76660508e-01 -6.38769388e-01 1.20743763e+00
5.95097840e-01 -3.90033066e-01 -4.11983013e-01 -3.55565369e-01
3.76996398e-01 -2.73954779e-01 -1.20724082e+00 1.91052246e+00
-3.08374941e-01 5.29453397e-01 1.95594281e-01 -7.69480288e-01
9.27139342e-01 1.66449025e-01 2.25879364e-02 -5.29186904e-01
-3.93453777e-01 1.65629253e-01 -4.84235048e-01 -3.86239648e-01
4.23293769e-01 6.92411587e-02 -2.99791783e-01 5.13593197e-01
3.86938363e-01 -2.69943267e-01 1.20012194e-01 3.82905960e-01
7.39755809e-01 -5.88396080e-02 1.76827729e-01 -1.85795411e-01
5.66488445e-01 -2.67550528e-01 2.78908074e-01 1.24104154e+00
-2.36120567e-01 9.38079119e-01 3.61620218e-01 -4.25809503e-01
-8.51510763e-01 -1.40354204e+00 -2.43889075e-02 1.50851488e+00
6.05907202e-01 -1.90839574e-01 -2.60634631e-01 -9.88967180e-01
3.06658655e-01 2.75846660e-01 -7.83630729e-01 1.56420991e-01
-3.17194283e-01 -3.29605327e-03 3.42655480e-01 5.70371151e-01
3.61187845e-01 -1.45086086e+00 -4.99353677e-01 -2.15485841e-02
-2.91450433e-02 -1.15866923e+00 -1.04421782e+00 -1.70179442e-01
-4.41825658e-01 -1.10995471e+00 -1.19575655e+00 -1.09554327e+00
1.02183235e+00 5.60408592e-01 1.27185357e+00 1.00161098e-01
-5.48305631e-01 1.01366258e+00 -2.59503782e-01 -4.42433953e-01
1.19019583e-01 -8.34078193e-02 -1.59759238e-01 3.63352776e-01
3.29780012e-01 -7.63751194e-02 -8.68350744e-01 5.75905979e-01
-1.06795013e+00 -2.92553812e-01 9.09737587e-01 1.01186097e+00
7.33023584e-01 -2.31911108e-01 6.43729806e-01 -5.21899819e-01
6.37527525e-01 -4.65004504e-01 -3.84992987e-01 7.14253724e-01
-5.22189820e-03 -6.35034731e-03 3.82680267e-01 -8.86390388e-01
-7.69110441e-01 2.78514624e-01 3.38147968e-01 -8.40332806e-01
-3.93518239e-01 -4.59739454e-02 -2.86568433e-01 -2.81474352e-01
4.62285459e-01 4.87477720e-01 -1.50870070e-01 -5.59451342e-01
5.69027185e-01 4.79733199e-01 5.85197926e-01 -9.81562555e-01
9.41814065e-01 6.57520831e-01 -2.98798800e-01 -9.09471929e-01
-5.92300951e-01 -7.64718115e-01 -5.83928227e-01 -4.37241271e-02
5.54496527e-01 -8.76650453e-01 -5.33680439e-01 1.45564526e-01
-1.35510874e+00 4.27258573e-02 -1.54208034e-01 1.28713921e-01
-5.31682312e-01 2.19690427e-01 -1.73951209e-01 -8.49060953e-01
-4.44520026e-01 -1.34381318e+00 1.96719623e+00 2.49461487e-01
-2.56535667e-03 -6.77192926e-01 -3.38304162e-01 -6.69162348e-02
3.86846632e-01 9.38582048e-02 7.91219413e-01 -7.52888262e-01
-1.01027048e+00 -5.39410353e-01 -6.86546922e-01 1.37812257e-01
2.14344129e-01 -2.97235221e-01 -8.73172343e-01 -5.63335598e-01
-4.01537091e-01 -7.57413447e-01 8.70045781e-01 7.18539231e-04
1.28504992e+00 -2.28593484e-01 -6.78110540e-01 3.14249098e-01
1.46803880e+00 7.84624647e-03 1.73135683e-01 -2.44470030e-01
5.06482780e-01 6.53583467e-01 8.44202042e-01 6.16592057e-02
3.19779329e-02 9.26339686e-01 3.61049116e-01 -3.66086841e-01
-1.35815978e-01 -4.99827504e-01 6.22361042e-02 -3.23715061e-02
3.71876329e-01 -1.69765949e-01 -6.98706150e-01 9.07364964e-01
-1.82391906e+00 -8.26034129e-01 4.75444734e-01 2.16396070e+00
4.79979455e-01 -2.33245268e-01 8.77592266e-02 -4.29644585e-01
6.91093922e-01 2.75454074e-01 -5.75864673e-01 2.17862830e-01
-9.39658284e-02 2.86472976e-01 2.20313668e-01 2.52370030e-01
-1.19696569e+00 1.13935685e+00 5.97175026e+00 1.07613206e+00
-1.18316102e+00 3.38290073e-02 3.95150870e-01 3.61779146e-02
-9.97799635e-02 -1.78863466e-01 -9.05959845e-01 4.16755050e-01
3.48007977e-02 1.00135781e-01 3.10977697e-01 1.08562016e+00
-2.60358274e-01 -1.39001712e-01 -1.53713131e+00 1.31091237e+00
3.55268925e-01 -1.17043698e+00 4.45060819e-01 -2.40908504e-01
4.21046913e-01 -1.05813123e-01 3.82718146e-01 4.29368705e-01
-2.72910353e-02 -1.06488943e+00 9.26853836e-01 6.20853603e-01
9.97063160e-01 -5.88928401e-01 4.72987384e-01 2.94666618e-01
-1.45195663e+00 -1.21034987e-01 -5.42629361e-01 5.22745490e-01
-1.50382623e-01 -8.10744390e-02 -8.19821894e-01 3.01123530e-01
6.43965006e-01 4.32564944e-01 -9.43623841e-01 1.14984429e+00
-2.51176894e-01 7.86896199e-02 -2.84659773e-01 -1.62479997e-01
3.79120409e-01 5.47488257e-02 4.74739045e-01 1.28781688e+00
1.18932001e-01 -7.04527870e-02 3.07483435e-01 1.50089693e+00
-9.17823166e-02 4.11555506e-02 -7.73544848e-01 -2.79823318e-02
6.41333520e-01 1.24275458e+00 -7.53177643e-01 -3.99827152e-01
-4.07625973e-01 1.29813433e+00 4.71393615e-01 9.41933692e-01
-6.34780288e-01 -6.29668653e-01 5.04798234e-01 5.28671313e-03
7.58733034e-01 -5.15314080e-02 4.67168361e-01 -1.01980221e+00
3.08460683e-01 -6.20931804e-01 2.28669018e-01 -9.56572533e-01
-1.65631473e+00 5.09227395e-01 1.50493637e-01 -1.03255725e+00
-5.42785674e-02 -7.42896020e-01 -7.04175532e-01 1.11848521e+00
-1.40682018e+00 -1.64880514e+00 -3.57518435e-01 6.04353607e-01
8.60759437e-01 -2.26120353e-01 7.42654383e-01 2.89430946e-01
-1.90622568e-01 6.57024443e-01 -1.60390243e-01 4.56903458e-01
8.66097987e-01 -1.28491938e+00 4.68988597e-01 4.97227043e-01
6.51062191e-01 8.73155177e-01 2.80116379e-01 -4.76591587e-01
-1.48206723e+00 -1.28706217e+00 6.76343083e-01 -8.32044125e-01
4.08148170e-01 -8.94280076e-01 -1.01932073e+00 5.51449537e-01
3.84187629e-03 7.01946676e-01 1.71862870e-01 -1.83334544e-01
-6.24287844e-01 -1.37384728e-01 -1.07480431e+00 4.74460274e-01
9.23360884e-01 -9.50723112e-01 -6.70209765e-01 4.49020833e-01
4.62516814e-01 -2.31803179e-01 -4.99883145e-01 2.62910396e-01
7.25344956e-01 -6.01286411e-01 1.40268457e+00 -7.06610322e-01
-2.63632387e-02 -5.18573761e-01 -3.69080693e-01 -8.99888217e-01
-1.71198398e-01 -2.66080052e-01 -5.05035073e-02 1.24576104e+00
1.02802522e-01 -3.13565284e-01 7.28084385e-01 2.90707767e-01
4.28247094e-01 -8.79402101e-01 -8.16394448e-01 -8.29684794e-01
-9.05063301e-02 -6.21653982e-02 5.75553000e-01 6.64230108e-01
-5.19938707e-01 2.99992859e-01 -1.74435556e-01 2.01036036e-01
6.36972249e-01 5.41598082e-01 1.05390453e+00 -1.07989359e+00
-1.34560093e-01 -5.73407114e-01 -4.74162310e-01 -1.47864854e+00
2.61353552e-01 -7.91558683e-01 2.18018696e-01 -1.25713170e+00
3.26356143e-01 -5.54403007e-01 -3.71169239e-01 4.22030658e-01
-3.66472602e-01 3.81292433e-01 3.95639509e-01 3.97434086e-01
-9.86832976e-01 5.68401575e-01 1.16379511e+00 -4.91269171e-01
8.32468197e-02 9.03406218e-02 -5.11664808e-01 3.61634523e-01
1.90806136e-01 -4.13065463e-01 -4.33028966e-01 -3.48890394e-01
-2.13613272e-01 1.45572707e-01 9.83193040e-01 -7.18250990e-01
3.32241386e-01 -6.50402680e-02 5.72608471e-01 -1.02758789e+00
5.75071037e-01 -9.32385921e-01 -1.44229904e-01 1.13715746e-01
-3.72472227e-01 2.15220582e-02 1.02233909e-01 9.37397718e-01
-3.15731078e-01 -2.33710304e-01 7.32552886e-01 -3.13560337e-01
-8.70094895e-01 5.30108333e-01 1.79721430e-01 5.98219149e-02
1.01052690e+00 -1.55600324e-01 -1.68679520e-01 -3.46422881e-01
-6.55300140e-01 2.05077901e-01 4.93648589e-01 7.57460773e-01
9.31569219e-01 -1.62298512e+00 -6.08126879e-01 2.57545590e-01
7.19609141e-01 1.44446880e-01 5.05926125e-02 4.87342626e-01
-1.20750606e-01 6.52781308e-01 6.17510267e-02 -9.58470583e-01
-1.10475707e+00 9.26588833e-01 2.97064245e-01 1.78174376e-01
-5.12036264e-01 1.13230312e+00 8.28128576e-01 -3.99055690e-01
5.93157232e-01 -3.22440892e-01 2.17195675e-01 -1.93689112e-02
6.40731215e-01 -4.28239768e-03 -2.97916919e-01 -8.23949158e-01
-4.59680915e-01 8.56280863e-01 -1.99251592e-01 -3.79972309e-02
9.49794412e-01 9.08810794e-02 3.64144295e-02 4.62593049e-01
1.41702902e+00 1.08197834e-02 -1.19498515e+00 -6.58678353e-01
4.02539559e-02 -7.28846073e-01 -3.49019289e-01 -8.55460882e-01
-8.99365246e-01 9.30454016e-01 6.48029864e-01 -1.36212349e-01
7.59551406e-01 5.26587963e-01 3.18233967e-01 5.66063046e-01
4.63605553e-01 -7.70546734e-01 6.13953531e-01 1.21862136e-01
1.53427052e+00 -1.52041924e+00 -9.73734409e-02 -9.29859430e-02
-5.27928174e-01 9.41203356e-01 6.73358917e-01 -3.36194843e-01
3.57554257e-01 -2.05788165e-01 -1.09388821e-01 -3.94501269e-01
-5.23120701e-01 -2.36994654e-01 7.49688148e-01 5.49485564e-01
2.28015836e-02 -1.16920881e-01 2.92390883e-01 4.29278195e-01
4.02551115e-01 -3.99421036e-01 -1.63703024e-01 9.16730583e-01
-3.95536721e-01 -9.56575453e-01 -6.15265429e-01 2.96501398e-01
-2.19740883e-01 8.43877867e-02 -8.15176189e-01 1.01314199e+00
2.78216638e-02 6.11126840e-01 -1.18057989e-02 1.87437713e-01
2.58498311e-01 1.24793939e-01 4.69086587e-01 -8.72480094e-01
-2.01145545e-01 1.80531219e-02 -5.77294469e-01 -6.16125584e-01
-2.07537830e-01 -4.04620618e-01 -8.18701208e-01 4.54181165e-01
-3.70886743e-01 1.96699172e-01 6.32277012e-01 6.84438169e-01
4.36216891e-01 1.53687879e-01 4.44459975e-01 -1.11277580e+00
-5.36965787e-01 -9.25280631e-01 -5.55575430e-01 6.48443103e-01
6.31187737e-01 -9.49331880e-01 -2.28293255e-01 -2.11332068e-01] | [11.648612022399902, 0.6234824061393738] |
83360a1f-1d71-4019-ae56-acac90a38986 | capturing-humans-in-motion-temporal-attentive | 2203.08534 | null | https://arxiv.org/abs/2203.08534v1 | https://arxiv.org/pdf/2203.08534v1.pdf | Capturing Humans in Motion: Temporal-Attentive 3D Human Pose and Shape Estimation from Monocular Video | Learning to capture human motion is essential to 3D human pose and shape estimation from monocular video. However, the existing methods mainly rely on recurrent or convolutional operation to model such temporal information, which limits the ability to capture non-local context relations of human motion. To address this problem, we propose a motion pose and shape network (MPS-Net) to effectively capture humans in motion to estimate accurate and temporally coherent 3D human pose and shape from a video. Specifically, we first propose a motion continuity attention (MoCA) module that leverages visual cues observed from human motion to adaptively recalibrate the range that needs attention in the sequence to better capture the motion continuity dependencies. Then, we develop a hierarchical attentive feature integration (HAFI) module to effectively combine adjacent past and future feature representations to strengthen temporal correlation and refine the feature representation of the current frame. By coupling the MoCA and HAFI modules, the proposed MPS-Net excels in estimating 3D human pose and shape in the video. Though conceptually simple, our MPS-Net not only outperforms the state-of-the-art methods on the 3DPW, MPI-INF-3DHP, and Human3.6M benchmark datasets, but also uses fewer network parameters. The video demos can be found at https://mps-net.github.io/MPS-Net/. | ['Hong-Yuan Mark Liao', 'Tyng-Luh Liu', 'Jen-Chun Lin', 'Wen-Li Wei'] | 2022-03-16 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Wei_Capturing_Humans_in_Motion_Temporal-Attentive_3D_Human_Pose_and_Shape_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Wei_Capturing_Humans_in_Motion_Temporal-Attentive_3D_Human_Pose_and_Shape_CVPR_2022_paper.pdf | cvpr-2022-1 | ['3d-human-pose-and-shape-estimation'] | ['computer-vision'] | [-5.25524616e-01 -5.84882736e-01 -9.61488560e-02 -2.05796152e-01
-7.58002251e-02 -2.36877650e-01 3.86130780e-01 -4.73611116e-01
-4.66097444e-01 3.66543889e-01 5.45802593e-01 2.21402600e-01
1.74812004e-01 -4.96167243e-01 -4.40097123e-01 -5.26502609e-01
-3.05517048e-01 1.74205944e-01 5.13768971e-01 -3.29241753e-01
-2.05197901e-01 4.70751375e-01 -1.41824889e+00 1.54142186e-01
4.67905998e-01 9.53453600e-01 4.20449883e-01 8.25990200e-01
3.48816991e-01 9.43592548e-01 -2.80192018e-01 1.12594210e-01
2.33296216e-01 -4.33626533e-01 -6.88148141e-01 2.20095560e-01
3.02574337e-01 -7.95059085e-01 -1.10883522e+00 6.26909435e-01
7.50629783e-01 4.85193163e-01 2.87986338e-01 -1.31607616e+00
-3.89539570e-01 3.87512445e-02 -6.00600123e-01 6.47753656e-01
7.11063921e-01 7.74616838e-01 5.41769624e-01 -9.49245811e-01
7.23691404e-01 1.56105852e+00 6.17271721e-01 5.55442393e-01
-5.57088315e-01 -6.20679915e-01 3.60669643e-01 6.31130695e-01
-1.51426184e+00 -4.27156270e-01 8.07101905e-01 -4.30980533e-01
9.59248960e-01 5.67137171e-03 1.29246330e+00 1.20122910e+00
2.39243299e-01 1.19668305e+00 3.85227591e-01 7.32102692e-02
-1.85606912e-01 -7.58095026e-01 -1.06200099e-01 8.97121906e-01
-5.54889552e-02 2.61701077e-01 -7.43684411e-01 1.68066770e-01
1.24206424e+00 2.66473919e-01 -5.95678866e-01 -3.16676289e-01
-1.46305025e+00 4.65456575e-01 6.78274572e-01 2.82942235e-01
-5.85753202e-01 5.49223423e-01 3.64295065e-01 -4.26407419e-02
2.69684643e-01 -1.00587659e-01 -3.84934425e-01 -2.46976286e-01
-8.51624131e-01 4.88799989e-01 3.30138952e-01 9.01596069e-01
5.84516585e-01 6.92498013e-02 -3.99540395e-01 4.74165529e-01
3.92646939e-01 6.01453364e-01 2.85569072e-01 -1.26406860e+00
4.33513492e-01 4.90645349e-01 2.88253933e-01 -1.26280892e+00
-7.10114360e-01 -2.46533513e-01 -8.22050810e-01 -1.97202742e-01
2.57187009e-01 -1.72933921e-01 -1.02233493e+00 1.66297674e+00
6.43144727e-01 3.81326735e-01 -3.23493034e-01 1.52874947e+00
9.44797218e-01 7.82365620e-01 1.45009920e-01 2.54451260e-02
1.33101666e+00 -1.22956908e+00 -5.71359873e-01 -4.41157490e-01
4.02219921e-01 -4.33297396e-01 7.64396548e-01 -6.25628754e-02
-1.05632365e+00 -9.03325260e-01 -8.40879023e-01 -2.59933263e-01
1.22991323e-01 1.62576005e-01 5.03122568e-01 -3.50221395e-02
-8.38472843e-01 3.55988711e-01 -1.29810369e+00 -4.41796362e-01
3.65816414e-01 2.57586211e-01 -4.62430596e-01 -2.97678143e-01
-1.43802416e+00 5.78348696e-01 3.05809826e-01 6.14180207e-01
-9.97879446e-01 -4.47445393e-01 -1.10727382e+00 -9.33809280e-02
4.20953929e-01 -1.30842483e+00 1.12726474e+00 -7.96246529e-01
-1.37618911e+00 4.88303006e-01 -2.83346862e-01 -3.45530391e-01
6.70143247e-01 -6.13478661e-01 -1.33125380e-01 5.21246493e-01
3.48318517e-02 1.08122385e+00 7.36629605e-01 -9.69108045e-01
-6.44634426e-01 -3.00347120e-01 1.66463237e-02 5.54487526e-01
-1.20665543e-02 -6.29263744e-02 -1.19365156e+00 -8.09560597e-01
5.25993779e-02 -9.79431987e-01 -2.72510409e-01 1.84662998e-01
-2.01522946e-01 -2.21359327e-01 7.49396980e-01 -8.27062666e-01
1.38618803e+00 -1.97736967e+00 5.75033784e-01 -1.05203658e-01
2.57441014e-01 3.74255002e-01 -7.33898580e-02 2.38391981e-01
2.78333008e-01 -3.26100916e-01 -5.59515469e-02 -4.67842519e-01
-1.54935926e-01 2.33769029e-01 1.44570142e-01 6.40931249e-01
1.36339307e-01 1.30431962e+00 -9.34835970e-01 -5.23021221e-01
5.62302411e-01 8.06942940e-01 -7.33337700e-01 4.31255758e-01
-1.33796096e-01 8.62092018e-01 -5.89398682e-01 7.60444105e-01
5.05550086e-01 -4.58756566e-01 -4.14693691e-02 -4.19517279e-01
-1.22478567e-01 -1.32374242e-01 -1.09953296e+00 2.06861329e+00
-9.31869901e-04 5.76561809e-01 -4.19408903e-02 -7.17642128e-01
5.23442328e-01 2.48291984e-01 8.58226955e-01 -5.58165610e-01
3.54390532e-01 -1.30018204e-01 -1.11463144e-01 -8.48087370e-01
4.98388380e-01 3.72211665e-01 -7.38051534e-02 -8.51513371e-02
6.46476373e-02 5.28154194e-01 7.96148032e-02 6.56885356e-02
1.10158932e+00 4.67233628e-01 2.90240854e-01 7.78696314e-02
7.08656371e-01 -8.12553763e-02 9.97986078e-01 4.06521648e-01
-6.75300896e-01 7.78770149e-01 1.31180257e-01 -8.24155629e-01
-9.70967233e-01 -1.02577269e+00 4.54958379e-01 9.75816131e-01
5.32854557e-01 -4.44022328e-01 -5.02647281e-01 -5.04026175e-01
-3.02883815e-02 -1.68277677e-02 -6.63560331e-01 -1.09094657e-01
-1.02618361e+00 -4.02775347e-01 4.30225760e-01 8.26588690e-01
9.01956558e-01 -1.19353044e+00 -1.01608837e+00 2.88738668e-01
-7.32492685e-01 -1.15685129e+00 -1.05681956e+00 -5.01536489e-01
-5.73528945e-01 -1.08829117e+00 -1.01246858e+00 -6.70771003e-01
3.96833956e-01 4.57868993e-01 9.31272030e-01 3.26935470e-01
-2.79372185e-01 5.16850948e-01 -4.77981091e-01 2.34688833e-01
2.63810962e-01 8.38961527e-02 1.01443708e-01 4.35896637e-03
3.65949452e-01 -5.59527457e-01 -1.21979964e+00 4.97756213e-01
-7.06481397e-01 3.45475227e-01 4.74026322e-01 7.62841105e-01
4.49951053e-01 -2.03849524e-01 2.01595262e-01 -1.89209372e-01
-3.03274039e-02 -6.49905205e-01 -2.23926082e-01 3.61080207e-02
5.08111641e-02 -1.83914721e-01 2.81109840e-01 -5.47043562e-01
-9.98120606e-01 4.06278521e-01 -2.05762729e-01 -9.47227180e-01
-2.54407495e-01 4.43220466e-01 -2.99166024e-01 1.99939162e-01
3.48056555e-01 2.72703469e-01 -1.79736659e-01 -2.77041435e-01
1.67213336e-01 1.88414052e-01 9.66070712e-01 -3.27319205e-01
6.95686042e-01 6.92110360e-01 -5.23312613e-02 -7.64236689e-01
-7.47347593e-01 -7.27179110e-01 -8.91548753e-01 -5.23726165e-01
1.24785900e+00 -1.34502852e+00 -8.64126742e-01 7.99765408e-01
-1.31467831e+00 -4.47413385e-01 2.15937138e-01 6.93820119e-01
-6.61645055e-01 5.27196646e-01 -8.98333371e-01 -6.79718256e-01
-3.91597778e-01 -1.00663829e+00 1.17547023e+00 2.55092084e-01
-2.11524576e-01 -8.11559975e-01 2.18811333e-02 3.03740382e-01
9.32996422e-02 4.86750275e-01 1.62570626e-01 1.41833663e-01
-7.09985733e-01 -5.11230379e-02 -9.96005908e-02 -3.05386465e-02
4.08227444e-02 -2.42347926e-01 -5.71207404e-01 -4.44331437e-01
-2.54964054e-01 -8.61712992e-02 1.08928561e+00 7.39088058e-01
8.26270223e-01 -2.94751585e-01 -3.40812236e-01 9.55968499e-01
8.22869837e-01 -5.51454199e-04 7.29221702e-01 3.38304102e-01
1.16603410e+00 5.31442285e-01 7.66864955e-01 7.19982207e-01
7.92580366e-01 9.14131939e-01 3.22905093e-01 1.02105914e-02
-2.38514140e-01 -4.76910323e-01 6.49358988e-01 7.99969316e-01
-3.98891538e-01 -2.01143667e-01 -9.21935380e-01 5.20019054e-01
-2.19332552e+00 -1.18394244e+00 6.58088624e-02 1.82776308e+00
5.04917741e-01 -2.89823413e-02 4.68916118e-01 -2.43669450e-01
7.48060226e-01 4.35712397e-01 -5.90787649e-01 5.83976924e-01
-6.73501492e-02 -3.13003123e-01 2.57487565e-01 5.16791463e-01
-1.25676012e+00 1.08382690e+00 4.61943483e+00 5.57687044e-01
-9.36696351e-01 6.28520772e-02 2.72922754e-01 -4.37819272e-01
6.28026798e-02 -1.27450228e-01 -7.42366016e-01 5.62106013e-01
4.46247548e-01 8.18766579e-02 2.88284719e-01 5.52415788e-01
6.46861255e-01 -7.24699423e-02 -9.79535282e-01 1.15969467e+00
5.84122017e-02 -1.12380576e+00 1.16464429e-01 3.78526784e-02
5.63958168e-01 8.14104155e-02 -1.32670477e-01 2.10603341e-01
-6.02932125e-02 -9.09996808e-01 9.63237941e-01 9.34906006e-01
4.75463659e-01 -9.45409834e-01 7.16012239e-01 4.15419251e-01
-1.90791643e+00 -2.16161802e-01 -2.51828104e-01 -1.38860628e-01
5.44081330e-01 2.75001109e-01 -2.62373447e-01 6.33279681e-01
9.87046123e-01 1.11978126e+00 -5.53338587e-01 1.17022896e+00
-2.20958993e-01 2.01683313e-01 -2.05555528e-01 2.54463196e-01
3.33362162e-01 2.23636970e-01 6.90144300e-01 1.15112329e+00
4.14115608e-01 5.63758612e-01 5.79900384e-01 5.76481044e-01
2.12139085e-01 -3.11465293e-01 -3.43528539e-01 2.00534210e-01
4.90903616e-01 1.07258177e+00 -5.73720038e-01 -3.88020843e-01
-4.25044596e-01 1.23964310e+00 3.13678980e-01 4.53185380e-01
-1.03495765e+00 -1.74181554e-02 8.20459545e-01 2.37937868e-01
5.42994022e-01 -7.59635985e-01 3.98701638e-01 -1.42843354e+00
-3.82005004e-03 -7.60664403e-01 6.38008714e-01 -8.88720334e-01
-1.10893333e+00 6.82742059e-01 1.67292818e-01 -1.41828001e+00
-4.17297244e-01 -3.27881962e-01 -5.13352752e-01 6.16289675e-01
-1.18997180e+00 -1.36180091e+00 -8.16356063e-01 9.80102658e-01
7.37025917e-01 1.25526428e-01 1.65115997e-01 3.83295417e-01
-5.94322085e-01 3.67187113e-01 -6.27625048e-01 4.35433656e-01
7.37842858e-01 -7.73043990e-01 7.62216687e-01 9.13636327e-01
-1.38467848e-01 4.67002988e-01 5.07507741e-01 -9.57260013e-01
-1.42411041e+00 -1.25369167e+00 6.22461855e-01 -5.16748250e-01
2.98997402e-01 -7.34579936e-02 -9.24995720e-01 6.46343410e-01
3.06278411e-02 3.77167702e-01 1.48858637e-01 -3.92280191e-01
-1.43912867e-01 1.39684424e-01 -6.61204040e-01 6.00678563e-01
1.58862960e+00 -3.39930445e-01 -5.74955583e-01 -5.70576563e-02
8.22800815e-01 -6.53873503e-01 -7.85406590e-01 5.51884294e-01
7.10132837e-01 -9.09992278e-01 1.26739633e+00 -3.75854939e-01
4.14013535e-01 -6.62597001e-01 -1.01963744e-01 -1.03871453e+00
-7.08561659e-01 -6.21130526e-01 -5.76873362e-01 7.56205797e-01
-3.18433851e-01 -1.10188246e-01 8.85799706e-01 3.81882906e-01
-1.83578327e-01 -7.58370519e-01 -9.21467960e-01 -6.15218163e-01
-2.93802887e-01 -3.96500260e-01 4.12088037e-01 7.37953782e-01
-2.34436363e-01 1.37720019e-01 -9.56437647e-01 2.71285355e-01
6.23792648e-01 -9.70035568e-02 1.00940502e+00 -8.64057004e-01
-3.47274691e-01 -2.76448280e-01 -6.01639628e-01 -1.68898118e+00
8.75798315e-02 -6.00437105e-01 1.93252578e-01 -1.41081893e+00
1.22899257e-01 4.59858254e-02 -1.16597928e-01 3.94850731e-01
-4.52382624e-01 2.15743423e-01 4.77885276e-01 4.16079164e-01
-9.46396172e-01 8.95436347e-01 1.51692367e+00 -1.18107339e-02
-3.50581437e-01 -1.16956629e-01 -6.80212453e-02 8.18797231e-01
6.47306204e-01 -2.08757728e-01 -3.19370955e-01 -6.04880929e-01
-1.26266450e-01 3.57829690e-01 1.00545239e+00 -1.32663357e+00
6.30436420e-01 -2.40711212e-01 9.19202447e-01 -1.03107512e+00
3.83402050e-01 -4.91730154e-01 3.42930913e-01 6.72961831e-01
-3.65629122e-02 4.15488183e-01 4.33622599e-02 7.43627071e-01
-1.27298564e-01 3.36427242e-01 4.89856869e-01 -4.15714532e-01
-1.30264688e+00 9.37922418e-01 -3.65508735e-01 1.56774357e-01
8.02866876e-01 -1.32080540e-01 -7.78140053e-02 -4.76408094e-01
-7.84932911e-01 7.05882847e-01 5.19055307e-01 6.56527877e-01
9.43627775e-01 -1.65586805e+00 -6.91868126e-01 1.34800598e-01
4.69446704e-02 2.40025535e-01 8.11653972e-01 9.96921837e-01
-6.89712942e-01 3.99027973e-01 -4.17437494e-01 -7.76267707e-01
-1.08346868e+00 4.88409489e-01 5.28942525e-01 -9.70174521e-02
-8.92643154e-01 6.84077382e-01 3.90483767e-01 -9.02052894e-02
3.09522867e-01 -1.82904527e-01 -1.55356511e-01 -2.21125484e-01
5.93582273e-01 5.99235415e-01 -4.94064003e-01 -1.10356784e+00
-6.19967818e-01 7.62332380e-01 2.55667955e-01 -1.21796109e-01
1.12679672e+00 -5.84495366e-01 1.96182281e-01 2.37608463e-01
1.21485674e+00 -3.88289005e-01 -1.99747300e+00 -3.01810890e-01
-3.90587538e-01 -6.20739579e-01 -1.43520355e-01 -2.95481712e-01
-1.25398469e+00 8.82126033e-01 5.06234944e-01 -4.27071393e-01
1.20057309e+00 1.53759299e-02 1.15044463e+00 3.04756969e-01
4.61518198e-01 -9.28547204e-01 4.10726190e-01 7.22720385e-01
1.03462839e+00 -1.11188650e+00 8.61832127e-02 -3.07209283e-01
-7.41231740e-01 1.10030639e+00 1.08479548e+00 -1.00844689e-01
6.00120544e-01 -2.63230894e-02 -5.01340777e-02 -2.41809785e-01
-7.93630481e-01 -4.58679467e-01 5.72296143e-01 5.39848864e-01
3.05616021e-01 -2.23580018e-01 -3.92483994e-02 7.24135518e-01
-1.27587272e-02 1.76605850e-01 1.49600908e-01 9.15716171e-01
-4.49075490e-01 -5.98404467e-01 -3.30476135e-01 -5.44678271e-02
-1.66173592e-01 1.53186440e-01 -1.44495919e-01 7.90508568e-01
3.86404812e-01 7.58598387e-01 1.23328447e-01 -7.72191644e-01
3.71407509e-01 -3.24140638e-01 4.87326652e-01 -2.64987290e-01
-4.02835965e-01 2.79385924e-01 -1.05565051e-02 -1.00201929e+00
-7.62740135e-01 -7.36441731e-01 -1.41197765e+00 -6.14362478e-01
1.60073936e-01 -1.75715655e-01 -1.83018774e-01 9.09474552e-01
4.37198967e-01 5.70006549e-01 4.24245328e-01 -1.48560727e+00
7.74632394e-03 -8.63277078e-01 -1.89819485e-01 4.60417509e-01
4.60559577e-01 -9.46631193e-01 -1.13347955e-01 1.38748035e-01] | [7.252886772155762, -0.5278599858283997] |
d50f3cfa-0787-4dd2-97dd-7cbe9cc66f4a | towards-automated-feature-engineering-for | 1909.01185 | null | https://arxiv.org/abs/1909.01185v1 | https://arxiv.org/pdf/1909.01185v1.pdf | Towards automated feature engineering for credit card fraud detection using multi-perspective HMMs | Machine learning and data mining techniques have been used extensively in order to detect credit card frauds. However, most studies consider credit card transactions as isolated events and not as a sequence of transactions. In this framework, we model a sequence of credit card transactions from three different perspectives, namely (i) The sequence contains or doesn't contain a fraud (ii) The sequence is obtained by fixing the card-holder or the payment terminal (iii) It is a sequence of spent amount or of elapsed time between the current and previous transactions. Combinations of the three binary perspectives give eight sets of sequences from the (training) set of transactions. Each one of these sequences is modelled with a Hidden Markov Model (HMM). Each HMM associates a likelihood to a transaction given its sequence of previous transactions. These likelihoods are used as additional features in a Random Forest classifier for fraud detection. Our multiple perspectives HMM-based approach offers automated feature engineering to model temporal correlations so as to improve the effectiveness of the classification task and allows for an increase in the detection of fraudulent transactions when combined with the state of the art expert based feature engineering strategy for credit card fraud detection. In extension to previous works, we show that this approach goes beyond ecommerce transactions and provides a robust feature engineering over different datasets, hyperparameters and classifiers. Moreover, we compare strategies to deal with structural missing values. | ['Liyun He-Guelton', 'Léa Laporte', 'Pierre-Edouard Portier', 'Yvan Lucas', 'Sylvie Calabretto', 'Olivier Caelen', 'Michael Granitzer'] | 2019-09-03 | null | null | null | null | ['automated-feature-engineering'] | ['methodology'] | [ 2.80722171e-01 -2.17131317e-01 -3.40329051e-01 -4.55874681e-01
9.31130257e-03 -2.84998924e-01 6.35909379e-01 4.69911277e-01
-4.99398291e-01 6.83845520e-01 -2.62795359e-01 -5.31614900e-01
-3.22226912e-01 -1.19068241e+00 -2.99000919e-01 -5.69253445e-01
-1.96278036e-01 8.56375933e-01 3.08765769e-01 -1.65552378e-01
6.08309090e-01 7.09836364e-01 -1.62461555e+00 7.16013014e-01
3.87281060e-01 9.52977479e-01 -2.46616080e-01 5.80231965e-01
-2.26922378e-01 7.98133492e-01 -6.40621901e-01 -5.54099560e-01
3.05040956e-01 -5.79853952e-01 -6.52193069e-01 7.86517933e-02
-3.43848825e-01 -3.90489370e-01 1.47854716e-01 5.74097037e-01
-3.15582454e-01 -2.18256980e-01 7.81021416e-01 -1.37742209e+00
6.69256374e-02 4.15421367e-01 -4.38301504e-01 1.06767416e-01
6.96770847e-01 -2.11233124e-01 8.32683623e-01 -5.42250335e-01
7.65788794e-01 6.68582618e-01 9.40446854e-01 2.57305056e-02
-1.21698439e+00 -3.43899816e-01 -1.48233831e-01 4.24239159e-01
-7.94297934e-01 -7.45241810e-03 5.66744983e-01 -5.39560139e-01
1.32853639e+00 2.22944841e-01 1.09889364e+00 9.74020183e-01
2.85173625e-01 6.28381073e-01 9.17012155e-01 -9.36708629e-01
4.51822519e-01 2.43488908e-01 5.19172668e-01 2.22852707e-01
6.28798068e-01 2.22300097e-01 -4.45539325e-01 -6.96608365e-01
5.71618915e-01 6.39930248e-01 2.44346812e-01 -2.92595088e-01
-9.75428164e-01 1.05910516e+00 -3.49427551e-01 5.87719023e-01
-6.48514748e-01 -1.11078702e-01 6.02198005e-01 6.93536818e-01
-1.42600596e-01 6.08023554e-02 -5.86675465e-01 -4.35356021e-01
-1.09511268e+00 6.05735183e-01 1.47414255e+00 6.27269328e-01
7.11419761e-01 -2.74681181e-01 2.28450596e-01 5.17804623e-01
2.56809056e-01 -2.80221775e-02 6.10805631e-01 -6.17403865e-01
6.07652485e-01 9.39346910e-01 4.54812348e-01 -9.36838984e-01
-2.73482502e-01 1.16115615e-01 -7.07405686e-01 3.65814716e-01
8.00833821e-01 1.77656978e-01 -6.24697924e-01 1.08758712e+00
-4.72106319e-03 -8.18075016e-02 -5.30247428e-02 3.37800235e-01
-1.94091439e-01 4.11011219e-01 -6.95039928e-02 -6.91024065e-01
1.31323767e+00 -2.49818936e-01 -7.24231899e-01 2.19953790e-01
7.53633797e-01 -6.65937603e-01 2.99890190e-01 1.04527998e+00
-6.60640597e-01 -2.63635695e-01 -1.03485346e+00 5.78715563e-01
-5.44249058e-01 5.93190119e-02 6.87058568e-01 9.95903015e-01
-3.91490608e-01 8.27197731e-01 -1.06314445e+00 -3.43327940e-01
8.84972140e-02 3.90707254e-01 -1.83039725e-01 8.60301107e-02
-1.12755656e+00 1.09248173e+00 4.51036900e-01 4.28231768e-02
-2.71314345e-02 -1.53233977e-02 -4.39695060e-01 1.12225816e-01
1.58482432e-01 -5.78456581e-01 9.93238568e-01 -1.19138956e+00
-1.05796647e+00 8.23370695e-01 -1.77383900e-01 -9.30690706e-01
1.07094383e+00 -4.86338027e-02 -5.52408814e-01 5.29365521e-03
-1.29220173e-01 -2.44690433e-01 8.38830352e-01 -8.39148223e-01
-8.87157440e-01 -6.52263880e-01 -3.97260129e-01 -2.90833533e-01
5.82758896e-02 2.52322555e-01 -1.93550363e-02 -7.06601262e-01
4.66664344e-01 -7.83135891e-01 -1.39078160e-03 -4.47031587e-01
-1.43270522e-01 3.66923027e-02 8.36081684e-01 -7.45830655e-01
1.56601167e+00 -1.69595432e+00 -3.25290501e-01 6.44953012e-01
-3.69282871e-01 1.91231519e-01 7.78130233e-01 1.05515134e+00
-2.61237174e-01 1.11717723e-01 -4.97713387e-01 -1.12805851e-02
-1.68664932e-01 3.97708058e-01 -3.18154126e-01 3.44083428e-01
5.58518469e-02 4.31546599e-01 -4.71202970e-01 -4.21050042e-01
4.11729217e-01 1.06953561e-01 -2.23357543e-01 2.62851194e-02
6.26313388e-02 -2.37443559e-02 -2.84888804e-01 5.36538661e-01
6.48276091e-01 3.31814378e-01 6.85920060e-01 2.96904594e-01
-4.83052656e-02 6.06356859e-02 -1.54229009e+00 8.75078440e-01
-1.71897635e-01 1.28235057e-01 -4.02560234e-01 -1.35232186e+00
1.28918934e+00 5.60239911e-01 6.72163188e-01 -4.99589622e-01
-1.43004030e-01 4.32437688e-01 -1.71035364e-01 -5.72034657e-01
3.33515167e-01 -4.68026727e-01 -6.89955354e-02 7.45202541e-01
-1.29374877e-01 4.36958551e-01 3.65667671e-01 -2.42188901e-01
1.17337930e+00 1.94563583e-01 9.99153018e-01 2.83644080e-01
5.07906497e-01 -2.76489276e-02 7.32189715e-01 9.34789777e-01
-1.01944387e-01 3.31549704e-01 9.39868927e-01 -7.86162376e-01
-1.05606043e+00 -6.88064814e-01 -2.66178548e-01 5.46751261e-01
-2.68589884e-01 -1.00860767e-01 -4.13499564e-01 -6.99734926e-01
4.37087744e-01 7.33830571e-01 -7.26789474e-01 1.80594772e-01
-7.17922032e-01 -1.04372680e+00 4.28900748e-01 3.58743608e-01
2.67787248e-01 -1.27597320e+00 -1.11495650e+00 7.25654364e-01
-1.52153280e-02 -6.25755131e-01 1.50015280e-01 5.32278657e-01
-1.49034739e+00 -1.31566179e+00 -2.40166321e-01 -3.78293931e-01
2.17677847e-01 -2.57836252e-01 1.00215042e+00 2.84549236e-01
-4.14730757e-01 -1.71516031e-01 -7.43027031e-01 -4.07294601e-01
-7.33282268e-01 -3.09328645e-01 -1.43583581e-01 3.56248051e-01
8.20648968e-01 -6.83078468e-01 -2.13519886e-01 3.98289382e-01
-9.98418570e-01 -3.62507582e-01 5.70025086e-01 1.04199135e+00
2.81191677e-01 2.29001403e-01 6.97061360e-01 -1.13887262e+00
5.56865335e-01 -7.37310410e-01 -4.92227942e-01 3.57064188e-01
-1.02070475e+00 -1.18795540e-02 4.25746590e-01 -3.60958129e-01
-1.05862021e+00 3.06275934e-01 1.15521468e-01 -1.16901688e-01
-5.71049273e-01 6.18463874e-01 -7.75000255e-04 4.70540076e-01
4.22085196e-01 2.83073455e-01 1.24617428e-01 -5.54178536e-01
-2.55384803e-01 8.01991522e-01 2.39468768e-01 -3.37000489e-01
1.49316147e-01 5.33228815e-01 1.43952101e-01 -6.71506405e-01
1.16464965e-01 -7.85966039e-01 -1.12877524e+00 -5.73110506e-02
3.06254447e-01 -1.90584511e-01 -8.94542933e-01 9.27553475e-01
-1.09995353e+00 2.55989552e-01 -1.23197176e-01 5.85453212e-01
-6.59764946e-01 6.33972883e-01 -7.08015442e-01 -1.54426110e+00
-6.67890310e-02 -5.77082515e-01 4.31017876e-01 -5.54025285e-02
-5.83906293e-01 -1.05807650e+00 2.13236615e-01 2.69614160e-01
-1.27094552e-01 5.72108328e-01 1.00897348e+00 -1.10291255e+00
-1.85874134e-01 -9.62316394e-01 2.77459145e-01 3.29652518e-01
1.09420054e-01 1.71887845e-01 -5.11381269e-01 4.46791266e-04
4.13362473e-01 2.09200203e-01 8.30119491e-01 2.01448336e-01
5.17060935e-01 -4.60197926e-01 -3.76982093e-01 1.39996246e-01
1.60298991e+00 8.41830850e-01 9.64469194e-01 9.65776563e-01
-5.05996682e-02 7.70713449e-01 9.41620648e-01 8.62592459e-01
-3.20872143e-02 9.67922270e-01 2.31936485e-01 2.30155915e-01
8.47694993e-01 -1.70372367e-01 3.57433826e-01 1.04199789e-01
-3.35873753e-01 2.09645964e-02 -1.02676725e+00 4.82643992e-01
-2.17836261e+00 -1.56428850e+00 -6.46011949e-01 2.60361671e+00
4.88734543e-01 5.09592295e-01 6.53978586e-01 9.87662554e-01
6.88435435e-01 -4.07682210e-01 -3.08622271e-01 -8.90958846e-01
1.36397153e-01 2.82339513e-01 4.66219246e-01 2.51608819e-01
-9.31313217e-01 2.60292619e-01 5.89019012e+00 3.57138634e-01
-5.58551013e-01 -2.45714545e-01 3.70035380e-01 9.22724083e-02
1.53119350e-02 3.47382545e-01 -7.62947798e-01 7.34866977e-01
8.73335004e-01 2.58438736e-01 4.94031198e-02 6.75275028e-01
2.89020717e-01 -6.96104527e-01 -1.26511252e+00 6.97362423e-01
-8.70769769e-02 -9.96338248e-01 8.34276527e-02 4.31035548e-01
2.36115962e-01 -5.97571790e-01 -4.86616403e-01 8.92887637e-02
2.53662132e-02 -8.93453121e-01 7.40382671e-01 9.78200734e-01
1.79146647e-01 -8.42789650e-01 1.25757694e+00 6.11788511e-01
-8.68859053e-01 -3.51397783e-01 2.63895124e-01 -4.93951082e-01
3.09898555e-01 7.19107985e-01 -9.77528691e-01 8.85873199e-01
6.10807002e-01 5.38554609e-01 -2.57150561e-01 1.07629299e+00
1.67953819e-01 7.53429890e-01 -3.79034668e-01 2.86126453e-02
-9.13972929e-02 -5.84636152e-01 3.40159804e-01 1.27584207e+00
3.18855643e-01 -9.58780125e-02 -2.21996352e-01 7.43966758e-01
6.40926778e-01 9.55092814e-03 -5.39568424e-01 2.22359642e-01
4.48235184e-01 6.87728286e-01 -9.19722795e-01 -3.83780241e-01
-5.43452680e-01 9.98793960e-01 -9.19219404e-02 2.39446864e-01
-4.31439042e-01 -6.02093279e-01 2.36212939e-01 4.09212917e-01
5.71701050e-01 1.99468583e-02 -7.17166424e-01 -1.11056936e+00
4.50958908e-01 -7.33680010e-01 7.05123425e-01 -1.99859098e-01
-1.21892214e+00 1.27864778e-01 -6.73449636e-02 -1.39372885e+00
-9.01823223e-01 -4.62166131e-01 -7.03757346e-01 8.55401874e-01
-9.39316273e-01 -7.34952867e-01 1.60631508e-01 4.21881735e-01
1.98204309e-01 -3.16497266e-01 9.61476803e-01 -6.64688721e-02
-2.71839142e-01 1.77643895e-01 2.21178174e-01 1.75185576e-01
3.42614353e-01 -1.20333159e+00 3.16080302e-01 5.97930670e-01
1.01437218e-01 6.89455569e-01 7.21082985e-01 -9.45478559e-01
-8.97386074e-01 -5.17394722e-01 1.54050958e+00 -5.36268473e-01
5.10350227e-01 -8.38927329e-02 -1.15309286e+00 7.25485444e-01
-3.68059844e-01 -5.36378503e-01 7.53725231e-01 2.30657578e-01
-2.57208467e-01 -1.05266765e-01 -1.41909707e+00 -1.41428739e-01
3.52475762e-01 -2.84007639e-01 -1.17938590e+00 -1.72276631e-01
-2.99005359e-01 1.23014830e-01 -8.97652388e-01 2.36598134e-01
1.04380894e+00 -1.64235187e+00 7.74571836e-01 -6.52147949e-01
2.59807646e-01 7.82966893e-03 7.43989944e-02 -5.87504923e-01
6.41336069e-02 -4.39386785e-01 -2.73526341e-01 1.12872732e+00
2.21259072e-01 -7.43881881e-01 9.34537828e-01 6.35691404e-01
4.82654929e-01 -8.69951844e-01 -1.12221420e+00 -8.46900225e-01
-2.16974124e-01 -5.78838170e-01 6.07772708e-01 1.07993627e+00
4.43623573e-01 -3.64289820e-01 -4.24899697e-01 -3.90317678e-01
5.63174248e-01 3.52422237e-01 5.43807447e-01 -1.52392316e+00
-5.70430934e-01 -4.86755818e-01 -7.62736320e-01 -2.79694706e-01
-3.75208616e-01 -4.06855613e-01 -4.32752490e-01 -1.14735818e+00
2.37802327e-01 -5.19510150e-01 -2.22608462e-01 4.40269321e-01
2.01223761e-01 -2.81940997e-01 1.07232317e-01 7.23679304e-01
2.75245309e-02 -2.43804827e-01 3.00053835e-01 3.25186402e-01
-4.71726239e-01 7.51808822e-01 -1.98806599e-02 7.15552151e-01
6.39062285e-01 -7.14898527e-01 8.00972134e-02 5.02316773e-01
4.65411782e-01 7.62255728e-01 4.27749455e-01 -5.18132508e-01
8.68126154e-02 -2.62310416e-01 4.62950557e-01 -6.28359258e-01
3.21463011e-02 -8.93593848e-01 6.73237383e-01 8.53471100e-01
-9.52698514e-02 -3.70222814e-02 -2.56768852e-01 8.07442427e-01
-4.73953515e-01 -8.27987075e-01 5.17102063e-01 -3.36183667e-01
-3.35552573e-01 -2.97080517e-01 -7.58784652e-01 -5.63804090e-01
1.09796810e+00 -9.44008946e-01 2.52633542e-01 -2.47473925e-01
-1.17767704e+00 -1.15427233e-01 5.06720901e-01 1.98449865e-01
4.75781381e-01 -8.79109442e-01 -5.63543558e-01 3.56118262e-01
9.21529531e-02 -4.97003317e-01 -6.58387020e-02 8.91899943e-01
-6.59116209e-01 5.31487703e-01 -5.16458452e-01 -3.21786612e-01
-1.43283451e+00 3.26006562e-01 2.69698203e-01 -8.44937563e-01
-3.83805871e-01 1.99744359e-01 -7.57026434e-01 -1.17989719e-01
1.10751703e-01 -3.50079477e-01 -5.06507397e-01 5.39799750e-01
2.44182870e-01 7.08955944e-01 4.20501918e-01 -4.07001585e-01
-3.85695368e-01 1.88906744e-01 -3.05547982e-01 -4.15153921e-01
1.49423504e+00 1.02651134e-01 -6.21358603e-02 8.27993095e-01
8.37747931e-01 -2.26845205e-01 -9.08611774e-01 1.36607498e-01
8.91426623e-01 -5.71532369e-01 -5.37417650e-01 -8.89070153e-01
-5.90378284e-01 5.53653836e-01 3.47725511e-01 6.69463754e-01
1.01964951e+00 -4.28712159e-01 4.39788371e-01 3.94044310e-01
6.73252165e-01 -1.24619007e+00 -4.76929694e-01 2.32180104e-01
5.94242275e-01 -9.45965946e-01 -1.06820073e-02 -1.93342134e-01
-7.66126394e-01 1.56091940e+00 -1.17323011e-01 -1.98964596e-01
6.05542481e-01 2.08064556e-01 -1.69074580e-01 -1.19430378e-01
-7.48555243e-01 1.85458750e-01 -3.14071774e-01 5.73694646e-01
3.95287067e-01 1.75199375e-01 -8.60954165e-01 9.48230922e-01
-1.02836996e-01 6.66591108e-01 7.06261933e-01 1.27724802e+00
-3.81206900e-01 -1.47750270e+00 -5.73874772e-01 7.75114596e-01
-5.40110588e-01 1.84758812e-01 -4.51383501e-01 1.09335887e+00
3.60477448e-01 8.82362664e-01 6.48212433e-02 -2.05498457e-01
5.13130188e-01 5.57140052e-01 3.98529381e-01 -4.77332324e-01
-9.94996428e-01 -8.01178291e-02 2.49159798e-01 -3.28240395e-01
-3.41055930e-01 -1.49532819e+00 -1.16842747e+00 -1.38292432e-01
-5.60472131e-01 1.88337922e-01 7.14847982e-01 1.04048896e+00
-1.12783700e-01 -7.46321678e-02 6.77858233e-01 -4.33583438e-01
-6.81762934e-01 -9.76406395e-01 -1.07474554e+00 6.66788757e-01
1.50542527e-01 -6.24086618e-01 -5.37889898e-01 4.05265957e-01] | [7.992142200469971, 4.9940361976623535] |
e830f463-677b-473e-9225-bfa9750a5f83 | 3d-human-pose-shape-and-texture-from-low | 2103.06498 | null | https://arxiv.org/abs/2103.06498v1 | https://arxiv.org/pdf/2103.06498v1.pdf | 3D Human Pose, Shape and Texture from Low-Resolution Images and Videos | 3D human pose and shape estimation from monocular images has been an active research area in computer vision. Existing deep learning methods for this task rely on high-resolution input, which however, is not always available in many scenarios such as video surveillance and sports broadcasting. Two common approaches to deal with low-resolution images are applying super-resolution techniques to the input, which may result in unpleasant artifacts, or simply training one model for each resolution, which is impractical in many realistic applications. To address the above issues, this paper proposes a novel algorithm called RSC-Net, which consists of a Resolution-aware network, a Self-supervision loss, and a Contrastive learning scheme. The proposed method is able to learn 3D body pose and shape across different resolutions with one single model. The self-supervision loss enforces scale-consistency of the output, and the contrastive learning scheme enforces scale-consistency of the deep features. We show that both these new losses provide robustness when learning in a weakly-supervised manner. Moreover, we extend the RSC-Net to handle low-resolution videos and apply it to reconstruct textured 3D pedestrians from low-resolution input. Extensive experiments demonstrate that the RSC-Net can achieve consistently better results than the state-of-the-art methods for challenging low-resolution images. | ['Fernando de la Torre', 'Laszlo A. Jeni', 'Francesc Moreno-Noguer', 'Hao Chen', 'Xiangyu Xu'] | 2021-03-11 | null | null | null | null | ['3d-human-pose-and-shape-estimation'] | ['computer-vision'] | [ 2.12136969e-01 -2.22357690e-01 -1.45943061e-01 -4.01135832e-01
-7.16802597e-01 -7.33680204e-02 3.99457633e-01 -2.27079064e-01
-6.68898761e-01 7.88094580e-01 1.25199020e-01 3.30568522e-01
9.57693383e-02 -7.85024285e-01 -9.82969284e-01 -6.55256689e-01
1.62716776e-01 3.78724813e-01 5.73614776e-01 -2.04545438e-01
-1.95070118e-01 4.44890559e-01 -1.67118716e+00 2.24448591e-01
7.09843636e-01 1.10111594e+00 2.42153272e-01 5.04407406e-01
2.71646619e-01 5.43845475e-01 -3.53315860e-01 -4.22762305e-01
5.01882017e-01 -2.14050874e-01 -5.66473067e-01 3.04466099e-01
7.66439021e-01 -5.53641617e-01 -4.45036799e-01 1.08638823e+00
5.69085956e-01 1.28557920e-01 3.62024337e-01 -8.26514959e-01
-3.29498172e-01 8.49245265e-02 -1.00853550e+00 2.34748706e-01
4.21107322e-01 -9.32774022e-02 6.23611212e-01 -7.09346890e-01
5.85598648e-01 1.62684155e+00 8.35646749e-01 6.05718076e-01
-1.26346791e+00 -5.40392399e-01 1.80736944e-01 2.68119574e-01
-1.29867804e+00 -2.86614716e-01 9.56644893e-01 -1.92996949e-01
3.69792044e-01 2.63269413e-02 6.45786762e-01 1.20330870e+00
8.29958171e-02 7.94664145e-01 1.37066901e+00 -2.02252507e-01
-4.39090421e-03 -1.31051168e-01 -3.06904882e-01 6.69570446e-01
4.10539299e-01 2.95038968e-01 -5.72177768e-01 1.18206725e-01
1.44588792e+00 1.62667390e-02 -3.09056669e-01 -7.33367383e-01
-9.27777469e-01 6.55748606e-01 6.25730455e-01 -4.46320027e-02
-2.55097419e-01 -9.81750339e-02 3.87802720e-01 1.39028907e-01
7.55143464e-01 -1.37954667e-01 -3.24390858e-01 1.53920084e-01
-6.87498093e-01 4.54656750e-01 3.19675744e-01 6.78558648e-01
4.70476151e-01 -5.27840108e-02 1.79481909e-01 1.02265918e+00
2.03409836e-01 4.82510865e-01 2.84358829e-01 -9.62413073e-01
5.19612134e-01 2.96371877e-01 2.55655050e-01 -1.08023286e+00
-5.88239968e-01 -5.89950264e-01 -1.12608588e+00 4.78184968e-01
6.44214809e-01 5.25382981e-02 -8.51934731e-01 1.82195330e+00
6.83054626e-01 -2.99622174e-02 -1.51015401e-01 1.56570673e+00
8.31884980e-01 3.81577790e-01 -1.72515646e-01 -1.64560646e-01
1.22627580e+00 -9.08454657e-01 -5.73764145e-01 -3.26095432e-01
-1.13703437e-01 -7.20863760e-01 9.08834159e-01 4.18220103e-01
-1.26514566e+00 -8.68869424e-01 -1.10313916e+00 -2.70667046e-01
7.38126114e-02 1.11595601e-01 2.86617905e-01 4.10784274e-01
-6.47764087e-01 6.40470684e-01 -9.98013318e-01 -2.08266228e-01
3.22595626e-01 2.70655721e-01 -5.72811902e-01 -3.41263235e-01
-1.13772690e+00 8.63268256e-01 1.90385476e-01 2.20118538e-01
-5.30845463e-01 -3.74294639e-01 -1.16160595e+00 -1.33184373e-01
7.35044479e-01 -7.41197824e-01 1.00660574e+00 -9.70165968e-01
-1.55831075e+00 9.94713068e-01 6.74395561e-02 -3.69100541e-01
9.82936561e-01 -5.00272095e-01 -1.96715072e-01 3.58729362e-01
1.28158733e-01 4.92816567e-01 1.01690876e+00 -1.32039809e+00
-5.98254204e-01 -6.99221194e-01 1.65422186e-01 4.52919900e-01
-3.52615230e-02 -1.52097434e-01 -7.43965983e-01 -7.65082598e-01
2.31847346e-01 -7.56068170e-01 -2.32449383e-01 4.24964219e-01
-8.59130472e-02 1.47843994e-02 6.28563523e-01 -8.46846819e-01
6.30313516e-01 -1.98454583e+00 4.14921224e-01 -1.61292106e-01
-9.79751907e-03 2.68664211e-01 2.74288971e-02 -3.13234329e-01
1.41997203e-01 -3.78191561e-01 -1.79272264e-01 -3.74138653e-01
-2.44424820e-01 2.78764307e-01 1.02236770e-01 7.59305060e-01
2.68543988e-01 5.12926698e-01 -8.55701685e-01 -6.91060483e-01
3.43313664e-01 9.91987646e-01 -5.69111347e-01 2.25170940e-01
1.62319914e-02 7.38771319e-01 -2.85174817e-01 4.74054366e-01
9.37752187e-01 -2.35554755e-01 1.07916966e-01 -4.98140991e-01
-8.80085304e-02 4.79044728e-02 -1.40889573e+00 1.73173857e+00
-4.05735403e-01 1.88531145e-01 4.82332736e-01 -1.08793104e+00
8.55038762e-01 1.25561550e-01 3.86719197e-01 -9.85759377e-01
6.50795624e-02 1.90517679e-01 -3.21967721e-01 -2.87186265e-01
2.80383706e-01 -4.06746417e-01 -5.40167093e-03 4.34358716e-02
-6.20633364e-02 2.13315398e-01 6.19076379e-02 -2.72924364e-01
6.31337643e-01 5.00880837e-01 2.63252497e-01 4.29832824e-02
7.07626641e-01 -4.45221275e-01 1.01473618e+00 4.62088019e-01
-1.71043530e-01 8.93423438e-01 1.25595301e-01 -7.47273564e-01
-1.11821997e+00 -1.15957165e+00 -1.74020618e-01 8.22295010e-01
4.95299131e-01 -8.06709602e-02 -6.46132112e-01 -6.51037514e-01
-6.15843274e-02 -1.63760349e-01 -4.81388867e-01 -6.73286095e-02
-9.18570280e-01 -5.69454551e-01 6.92238361e-02 7.54161000e-01
1.01324081e+00 -7.83919036e-01 -7.97384977e-01 1.29353344e-01
-5.40273368e-01 -1.44658291e+00 -5.58012247e-01 -3.58130708e-02
-8.41604412e-01 -1.01640439e+00 -1.09976768e+00 -7.77177215e-01
6.10669732e-01 2.65648574e-01 1.01548409e+00 -8.41933489e-02
-3.07746530e-01 5.19423708e-02 -1.65330321e-01 3.21198963e-02
-1.15962900e-01 -2.92901695e-02 2.32608989e-01 1.94286302e-01
7.89359957e-03 -5.47243655e-01 -6.77549243e-01 4.64485139e-01
-6.99194849e-01 1.67945191e-01 6.99995637e-01 9.17366922e-01
9.78898406e-01 1.99041203e-01 3.59198958e-01 -6.39203548e-01
-3.81000410e-03 2.30274480e-02 -8.08177650e-01 -1.32929971e-02
-1.81255713e-01 8.12628716e-02 7.91040838e-01 -5.19778609e-01
-1.20186210e+00 3.66839796e-01 -3.50415707e-01 -6.65263414e-01
-1.79846197e-01 4.80028838e-02 -4.21461284e-01 -2.15900660e-01
5.76909721e-01 1.53020814e-01 1.64491475e-01 -8.19166958e-01
1.24549642e-01 3.81695539e-01 9.13514376e-01 -5.10276914e-01
1.01531541e+00 7.60437012e-01 2.36262709e-01 -8.79476786e-01
-1.28950274e+00 -4.70266044e-01 -8.10949624e-01 -1.81083933e-01
8.99070442e-01 -1.20932627e+00 -5.99948525e-01 5.40255070e-01
-9.26725030e-01 -1.47669271e-01 -9.00016651e-02 6.74462259e-01
-7.11950421e-01 5.55601716e-01 -7.31670141e-01 -7.54145443e-01
-2.21686751e-01 -1.03803933e+00 1.23508394e+00 3.36217314e-01
2.33399451e-01 -7.58741915e-01 -5.12316972e-02 6.21971130e-01
2.17689067e-01 5.64787567e-01 5.50838947e-01 4.23253849e-02
-4.99499410e-01 1.23443536e-03 -4.28250551e-01 3.99175167e-01
9.24600139e-02 -6.46982431e-01 -7.96826422e-01 -4.92346674e-01
1.11076031e-02 -6.44163847e-01 8.86198938e-01 5.58753252e-01
1.14727545e+00 -2.88959235e-01 -8.77986476e-03 7.84474134e-01
1.37090170e+00 -3.24119508e-01 4.72713381e-01 4.83540744e-01
8.13070297e-01 6.21236622e-01 7.29481101e-01 3.82526428e-01
4.87403661e-01 1.20925367e+00 4.84658867e-01 -3.59033376e-01
-2.27126911e-01 -2.35489681e-01 2.10830986e-01 4.51550186e-01
-4.85336035e-01 3.86094719e-01 -4.12328243e-01 3.12318474e-01
-1.90727580e+00 -1.03428495e+00 2.42039964e-01 2.48431945e+00
9.13661301e-01 2.60265827e-01 5.96792579e-01 2.65128285e-01
7.95255125e-01 1.85064688e-01 -7.34073162e-01 -1.31631419e-02
-1.29338250e-01 6.00680262e-02 3.83503377e-01 3.36337477e-01
-1.42639971e+00 6.61964417e-01 5.52005482e+00 7.17637718e-01
-1.16743135e+00 1.62325650e-01 5.31065583e-01 -2.19911337e-01
1.73984617e-01 -4.98688519e-01 -8.43561709e-01 3.98877800e-01
3.54709506e-01 2.47579217e-01 3.24703962e-01 8.60956311e-01
1.48239389e-01 1.98693597e-03 -8.88004184e-01 1.33292508e+00
3.70744705e-01 -9.77808774e-01 -1.47353366e-01 1.60988083e-03
5.70276082e-01 -2.83784539e-01 1.21950069e-02 1.21447004e-01
-6.02818169e-02 -9.37914252e-01 8.27896476e-01 3.63742113e-01
9.10414934e-01 -1.01480269e+00 7.50626028e-01 4.20376986e-01
-1.48353970e+00 -1.18961602e-01 -6.91320360e-01 9.25027952e-03
2.37098053e-01 5.49965084e-01 -1.10917114e-01 6.77090228e-01
1.12811220e+00 7.44376183e-01 -3.27393740e-01 1.06368470e+00
-1.23502947e-01 3.15811597e-02 -2.94119745e-01 3.98383230e-01
-1.33966237e-01 -9.15433317e-02 5.79484582e-01 1.00279975e+00
3.57669592e-02 1.31007761e-01 4.83292311e-01 7.07587361e-01
-7.34875277e-02 -4.03785929e-02 -2.98051953e-01 6.20741665e-01
5.22056688e-03 1.18772018e+00 -5.85288107e-01 -1.09877981e-01
-5.41736841e-01 1.15769112e+00 4.77350652e-01 8.24207887e-02
-7.07986593e-01 -4.57373969e-02 5.49536228e-01 6.08208299e-01
5.28162777e-01 -2.07799330e-01 -1.43787071e-01 -1.37159741e+00
4.02764708e-01 -9.90281105e-01 4.11219776e-01 -5.52638471e-01
-1.40886784e+00 5.65438509e-01 7.76955336e-02 -1.34003854e+00
-2.44768918e-01 -5.76183796e-01 -1.39984265e-01 7.00958014e-01
-1.86311591e+00 -1.15977073e+00 -5.10817468e-01 7.72352040e-01
6.34077132e-01 8.59414414e-02 3.81996572e-01 4.73086298e-01
-4.27461386e-01 5.75991571e-01 -2.21867651e-01 2.34446093e-01
8.34718049e-01 -1.21288371e+00 1.65982932e-01 7.42210805e-01
-1.37436256e-01 2.93760955e-01 7.46336102e-01 -4.75736499e-01
-1.31612027e+00 -1.13532281e+00 5.36857486e-01 -1.17661944e-02
6.14138469e-02 -3.57496768e-01 -1.04353547e+00 3.73380601e-01
-2.60373801e-01 5.32959223e-01 1.89293697e-01 2.64978688e-02
-4.37083423e-01 -3.75509769e-01 -1.22988653e+00 4.18905079e-01
1.15202618e+00 -3.01300496e-01 -5.60008109e-01 1.41158581e-01
4.87296313e-01 -7.83841789e-01 -8.68861079e-01 6.15449905e-01
7.28247702e-01 -1.08520401e+00 1.38624752e+00 -2.84320265e-01
4.29960161e-01 -3.88326854e-01 -1.00302920e-01 -1.09607780e+00
-4.25202876e-01 -2.15862200e-01 -2.33164281e-01 8.71554732e-01
-1.04418881e-01 -3.10228944e-01 7.73086727e-01 2.59630978e-01
1.37107730e-01 -6.78977072e-01 -1.10566640e+00 -9.08999979e-01
9.14967731e-02 5.59780607e-03 2.41766319e-01 5.80749094e-01
-4.21310455e-01 4.03635144e-01 -8.29018354e-01 3.30213785e-01
1.27720964e+00 3.02514166e-01 8.38081896e-01 -1.37778163e+00
-5.30694783e-01 -1.04508005e-01 -5.43681502e-01 -1.27956855e+00
6.17826208e-02 -4.15153086e-01 1.48255989e-01 -1.43803787e+00
3.67973179e-01 -2.19388917e-01 -4.17641103e-02 2.49205425e-01
-1.37706682e-01 5.88072479e-01 1.73502728e-01 2.82973945e-02
-6.26276255e-01 6.49983704e-01 1.43436968e+00 -1.94516592e-02
-1.22125454e-01 1.97478190e-01 -4.21155542e-01 1.07430303e+00
5.31529546e-01 -1.44682392e-01 -2.19844475e-01 -3.67053032e-01
6.36100471e-02 2.75668770e-01 6.55306160e-01 -1.12158012e+00
1.09750271e-01 -6.96919635e-02 9.25198317e-01 -6.98516011e-01
5.70851803e-01 -8.07648957e-01 -1.07639581e-01 4.54016507e-01
-1.99514285e-01 -2.51680352e-02 2.93531492e-02 6.73569322e-01
-1.68077856e-01 1.99955061e-01 1.35355878e+00 -3.78820956e-01
-4.82290179e-01 4.34105843e-01 7.92556256e-02 2.79286414e-01
6.04209363e-01 -1.71874791e-01 -7.01597705e-02 -4.31750357e-01
-7.37364948e-01 1.18843049e-01 6.29462302e-01 4.38482046e-01
8.22683930e-01 -1.50304079e+00 -8.16627502e-01 3.02130938e-01
-1.54219344e-01 3.78231108e-01 3.71286362e-01 7.19769955e-01
-3.43457013e-01 1.73855469e-01 -5.45447946e-01 -7.76152849e-01
-1.45675027e+00 6.02335572e-01 4.44685549e-01 -4.22229350e-01
-9.51181710e-01 4.02639568e-01 3.79439294e-01 -4.28599566e-01
4.01374429e-01 3.41172628e-02 -3.37215900e-01 -3.08669925e-01
7.07416117e-01 3.30682427e-01 -2.24020574e-02 -9.85389590e-01
-3.53385180e-01 1.10036433e+00 -1.90862194e-01 -7.24200681e-02
1.38576281e+00 -3.39526713e-01 2.88630903e-01 3.30854297e-01
1.12313986e+00 -2.55271882e-01 -1.74978065e+00 -6.21669352e-01
-4.83128488e-01 -6.93052173e-01 9.30679888e-02 -5.31222820e-01
-1.30095923e+00 8.66877735e-01 7.69609749e-01 -2.35888913e-01
1.37045479e+00 -1.53246492e-01 9.13272202e-01 1.27863616e-01
5.67758322e-01 -1.27374423e+00 2.78245151e-01 2.87657738e-01
8.62897038e-01 -1.50899792e+00 1.86100423e-01 -6.27437055e-01
-4.06765878e-01 1.05387187e+00 8.66324961e-01 -2.28318349e-01
4.98033911e-01 1.22035153e-01 -1.01062752e-01 1.82822898e-01
-5.23952425e-01 -3.42672378e-01 4.90569949e-01 6.71091795e-01
3.07527125e-01 -1.33316815e-01 -4.22832042e-01 6.06383204e-01
-3.99116380e-03 -1.92620733e-03 3.05418789e-01 7.77784169e-01
-3.13058227e-01 -8.92654240e-01 -7.26464450e-01 1.08365081e-01
-6.62148595e-01 4.03254092e-01 -1.61058366e-01 7.32422769e-01
1.67472035e-01 6.82579160e-01 3.95628624e-02 -1.86944485e-01
5.60181141e-01 -4.13976967e-01 8.22047412e-01 -3.31179380e-01
-1.47782400e-01 5.42100310e-01 -4.46201041e-02 -7.13302433e-01
-7.42028058e-01 -6.38930142e-01 -9.68794465e-01 -2.28690088e-01
-1.39031589e-01 -2.57538110e-01 4.40294087e-01 8.28690886e-01
2.86668874e-02 4.10681814e-01 5.92243254e-01 -1.19986284e+00
-6.64515436e-01 -6.79594338e-01 -5.62467694e-01 5.39806783e-01
5.15095234e-01 -1.05100119e+00 -1.87920317e-01 1.17841223e-02] | [7.113180160522461, -0.958512008190155] |
d5b269c9-42a0-467d-ae92-b327b5d6cc59 | branching-model-with-state-dependent | 2306.02893 | null | https://arxiv.org/abs/2306.02893v1 | https://arxiv.org/pdf/2306.02893v1.pdf | Branching model with state dependent offspring distribution for Chlamydia spread | Chlamydiae are bacteria with an interesting unusual developmental cycle. A single bacterium in its infectious form (elementary body, EB) enters the host cell, where it converts into its dividing form (reticulate body, RB), and divides by binary fission. Since only the EB form is infectious, before the host cell dies, RBs start to convert into EBs. After the host cell dies RBs do not survive. We model the population growth by a 2-type discrete-time branching process, where the probability of duplication depends on the state. Maximizing the EB production leads to a stochastic optimization problem. Simulation study shows that our novel model is able to reproduce the main features of the development of the population. | ['Máté Szalai', 'Péter Kevei'] | 2023-06-05 | null | null | null | null | ['stochastic-optimization'] | ['methodology'] | [ 2.50026137e-01 2.13115439e-01 4.62212786e-02 5.83265841e-01
6.97625399e-01 -6.09485805e-01 5.74135303e-01 1.45975947e-01
-5.11036992e-01 1.11775529e+00 -3.89563799e-01 -2.47757673e-01
3.00295889e-01 -9.86529529e-01 -5.33104002e-01 -1.38849998e+00
-1.53527990e-01 1.08137441e+00 3.21733087e-01 8.38504173e-03
4.39089760e-02 4.16679054e-01 -1.31350839e+00 -2.57822543e-01
6.09552920e-01 3.18159550e-01 4.48491067e-01 1.16284227e+00
-1.48493096e-01 4.44248766e-01 -7.22437561e-01 9.16687492e-03
2.08473027e-01 -6.14852726e-01 -4.98156369e-01 -2.48138636e-01
-9.75872159e-01 -1.20409079e-01 1.86649889e-01 6.57572925e-01
7.00782612e-02 -1.54230952e-01 1.26661432e+00 -1.17352617e+00
-1.19056568e-01 -2.72597913e-02 -6.20979548e-01 -2.09025759e-02
1.78890556e-01 5.45139574e-02 3.72641563e-01 -3.21614861e-01
1.20002151e+00 1.17683220e+00 3.48644078e-01 8.79068077e-01
-1.17496395e+00 -4.35541309e-02 -4.78942126e-01 -5.16188920e-01
-1.33911860e+00 -2.65436739e-01 -1.91552490e-02 -7.06880152e-01
8.11598241e-01 1.87706605e-01 1.31204247e+00 9.13110375e-01
8.57522905e-01 4.50796366e-01 1.33473110e+00 -4.04723167e-01
8.12985361e-01 -3.02992970e-01 3.16182077e-02 5.14140069e-01
1.23509848e+00 1.84152741e-02 -2.86805518e-02 -2.61419058e-01
9.92365420e-01 1.60448495e-02 -1.28393874e-01 -3.33694786e-01
-8.07561278e-01 5.78036606e-01 -4.63354230e-01 5.23530722e-01
-5.75723171e-01 3.02326024e-01 -1.90802738e-01 1.17740318e-01
-6.95700347e-02 2.23190412e-01 -3.13800126e-01 -3.93607289e-01
-5.21855593e-01 2.02018708e-01 1.30869436e+00 3.77023458e-01
4.79143828e-01 -4.83768970e-01 1.05854422e-01 3.43933523e-01
4.35016006e-01 6.60117447e-01 1.90514341e-01 -7.88608909e-01
-1.90395072e-01 7.09569573e-01 6.36356056e-01 -4.29804116e-01
-3.86382312e-01 -3.84528220e-01 -8.54878843e-01 3.67778122e-01
9.82885659e-01 -3.79346818e-01 -7.70434797e-01 1.59081924e+00
3.62273395e-01 -3.49683315e-02 3.16702038e-01 3.59585583e-01
3.02048117e-01 1.07849646e+00 -1.08125374e-01 -9.92802918e-01
1.42363763e+00 -4.60337311e-01 -5.77062845e-01 2.29169562e-01
3.21884096e-01 -5.53132057e-01 1.02801360e-01 5.35603285e-01
-1.43644941e+00 1.95198908e-01 -8.54756951e-01 5.95620155e-01
-4.31620628e-02 -5.10027587e-01 2.20119789e-01 7.54078150e-01
-1.05645990e+00 1.71431869e-01 -1.27510059e+00 -7.95159161e-01
-1.06415786e-01 2.48747826e-01 -8.25126916e-02 -1.30986487e-02
-7.77593017e-01 6.60264134e-01 -1.24682365e-02 -1.87183797e-01
-1.19050109e+00 6.68126391e-03 -2.33751938e-01 -2.08824594e-02
1.68792456e-01 -1.43490553e+00 9.37448204e-01 -2.92064339e-01
-1.64189732e+00 9.34273362e-01 -5.04220068e-01 -2.51800805e-01
5.73680758e-01 2.90862203e-01 2.55631238e-01 2.19323605e-01
-5.99659570e-02 3.75094146e-01 5.54829538e-01 -1.67666674e+00
-7.19026864e-01 -4.26916450e-01 -1.69276133e-01 4.79816161e-02
1.44634664e-01 9.16254595e-02 -5.81861317e-01 -2.13749543e-01
9.31098834e-02 -1.15422773e+00 -4.09887910e-01 -6.24374151e-01
-3.63840550e-01 -1.92063570e-01 5.50603151e-01 -3.59909534e-01
9.30690765e-01 -1.95550621e+00 4.40541565e-01 -1.11106664e-01
4.21954036e-01 2.90764868e-02 3.89261484e-01 8.77410650e-01
4.69740868e-01 3.76492292e-01 -4.13817585e-01 -2.97019392e-01
-4.51267689e-01 5.23276806e-01 2.92651027e-01 5.60191691e-01
5.98629005e-02 2.54984200e-01 -1.04008031e+00 -6.11919880e-01
-4.86191630e-01 3.91315579e-01 -7.02404797e-01 2.25505322e-01
-1.07721962e-01 8.13979447e-01 -6.04700208e-01 6.14812553e-01
7.97887743e-01 -4.04903978e-01 2.47730255e-01 6.18966699e-01
-6.18978798e-01 -2.08027631e-01 -1.04305315e+00 3.26929808e-01
-2.83855140e-01 2.02715725e-01 5.44742048e-01 -8.08697283e-01
7.11553931e-01 4.16039288e-01 5.83458483e-01 2.34239638e-01
3.78732890e-01 5.13181150e-01 3.14616591e-01 -3.47662419e-01
3.83563548e-01 -3.42799991e-01 1.05590373e-01 5.93808174e-01
-3.53525072e-01 -3.60531926e-01 7.34371424e-01 2.00718060e-01
1.39654076e+00 -8.45468715e-02 7.10430086e-01 -2.63034940e-01
4.72938061e-01 4.53597493e-02 9.25597727e-01 6.23800159e-01
-1.97522029e-01 3.54021102e-01 9.78932977e-01 -4.56286184e-02
-1.14632154e+00 -1.30458140e+00 -2.86637992e-01 3.82755995e-01
5.43595910e-01 1.58421069e-01 -1.09373546e+00 2.86548972e-01
5.28790383e-03 3.05558890e-01 -5.33998489e-01 8.18319917e-02
-7.14915335e-01 -1.19537067e+00 2.77151972e-01 -3.52682501e-01
4.69500422e-01 -9.66012478e-01 -1.01722038e+00 3.89945894e-01
-2.08897367e-01 -4.63882089e-01 1.59908012e-01 3.21169078e-01
-1.15567219e+00 -9.86223757e-01 -8.55940402e-01 -4.70845073e-01
7.96572149e-01 -7.00655952e-02 7.46035755e-01 4.68121707e-01
-5.25001705e-01 1.53110087e-01 -2.03730449e-01 -1.84336036e-01
-1.10475814e+00 -3.37156475e-01 3.58755849e-02 -3.27078104e-01
1.94758736e-02 -4.30564791e-01 -5.27034760e-01 3.42678756e-01
-7.36661077e-01 -1.41221061e-01 2.25678518e-01 7.11710215e-01
6.37887359e-01 4.40070778e-01 3.06931078e-01 -1.88339248e-01
3.88800889e-01 -6.98881626e-01 -8.01727772e-01 1.00883611e-01
-1.04856230e-01 -8.71387497e-02 6.60434484e-01 -1.43770620e-01
-1.31557846e+00 -5.34989461e-02 3.27514857e-01 4.65560287e-01
-1.89336371e-02 1.76091447e-01 -3.87139968e-03 2.81452656e-01
9.74870995e-02 4.27776903e-01 1.75424561e-01 -6.54101431e-01
-3.24970990e-01 6.97558343e-01 3.18197548e-01 -4.31090534e-01
3.77513409e-01 9.15884316e-01 5.77816248e-01 -1.11149287e+00
2.05246791e-01 -3.59307557e-01 -2.61333227e-01 -3.09284806e-01
1.00241458e+00 -5.89286566e-01 -1.08254242e+00 1.03579926e+00
-1.40389502e+00 -3.81124347e-01 -4.70459819e-01 4.79249269e-01
-7.62479246e-01 2.86431402e-01 -1.07099569e+00 -1.36735082e+00
1.91490889e-01 -8.68899882e-01 8.18434477e-01 6.63148582e-01
8.71814340e-02 -9.53857303e-01 8.22394907e-01 1.28292918e-01
2.14975148e-01 2.99825311e-01 8.63733768e-01 -4.40559745e-01
-9.34144735e-01 -3.96112278e-02 2.26268873e-01 -1.80487663e-01
5.73087111e-02 8.84173453e-01 -3.83285321e-02 -3.50827724e-01
2.88703471e-01 4.04321760e-01 7.84531593e-01 7.83993781e-01
8.14839092e-04 -1.54857531e-01 -1.02318716e+00 6.53703451e-01
1.58824217e+00 9.72749472e-01 5.02892137e-01 3.90492409e-01
-1.06401771e-01 6.15013301e-01 5.55898428e-01 6.25065207e-01
1.83521718e-01 3.71764034e-01 5.72165966e-01 2.54198998e-01
2.11765081e-01 3.71082574e-01 5.02719939e-01 9.22019601e-01
-9.56458151e-01 -7.70575345e-01 -1.06964612e+00 4.65302348e-01
-1.35751677e+00 -9.51958060e-01 -4.06654239e-01 2.38114071e+00
1.01884377e+00 3.22972722e-02 1.16060905e-01 -1.65689453e-01
8.23661327e-01 -4.96980071e-01 -1.95521623e-01 -6.40096366e-01
-3.11770052e-01 2.47043088e-01 7.01358914e-01 5.99530458e-01
-4.89439875e-01 5.18188059e-01 7.55720949e+00 3.12797338e-01
-9.84804034e-01 1.49046138e-01 6.14193201e-01 -1.71883516e-02
-2.03479946e-01 2.28349224e-01 -1.12399316e+00 6.54185653e-01
6.52834535e-01 -4.73160535e-01 4.19295847e-01 1.79979458e-01
3.86753529e-01 -9.25657511e-01 -8.05168509e-01 3.07535499e-01
-5.10388494e-01 -1.12741494e+00 -2.45648697e-01 7.26448953e-01
9.77372885e-01 -1.58872426e-01 -3.28052133e-01 -1.71472486e-02
8.12321961e-01 -7.08724737e-01 3.03319305e-01 6.94979429e-01
5.18246830e-01 -5.33251643e-01 7.95768082e-01 7.88643599e-01
-1.04674852e+00 -2.50382215e-01 -3.02628458e-01 -4.90384489e-01
6.87952876e-01 7.79881895e-01 -1.04928064e+00 2.83143997e-01
4.48369533e-01 -1.42995611e-01 -1.34822382e-02 1.22675681e+00
-3.45844813e-02 4.21572566e-01 -5.99484742e-01 -3.19395453e-01
-1.24082051e-01 -9.78585064e-01 1.15756667e+00 7.52714157e-01
5.98592401e-01 4.06801343e-01 -3.20434004e-01 8.12412083e-01
2.98228413e-01 1.29075423e-01 -5.73153555e-01 -2.17581719e-01
5.17336965e-01 8.11301529e-01 -1.28392184e+00 -5.72246134e-01
-6.04687084e-04 7.61354208e-01 -1.90806448e-01 3.40968519e-01
-2.63142288e-01 -2.71940917e-01 5.41413844e-01 3.12935233e-01
3.29551846e-01 -3.50460142e-01 -2.48443231e-01 -7.35492945e-01
-5.00246584e-01 -3.40811223e-01 1.05343029e-01 -3.95983428e-01
-6.36436403e-01 2.76668549e-01 8.54190364e-02 -6.59094632e-01
-4.88081306e-01 -4.61851358e-01 -8.44024897e-01 6.87599361e-01
-6.43508673e-01 -4.50685322e-01 -2.83942576e-02 1.51312709e-01
-7.09089497e-03 4.77991216e-02 4.57102269e-01 -1.12522878e-01
-7.57957220e-01 -1.92882717e-01 8.84965360e-01 -6.23542309e-01
5.16104698e-02 -1.07541311e+00 1.33774653e-01 7.12412775e-01
-7.68471897e-01 1.08199382e+00 1.16562271e+00 -1.23466289e+00
-1.27874017e+00 -5.53210497e-01 1.27346194e+00 -1.07168511e-01
4.35894340e-01 -1.46644861e-01 -3.57827008e-01 1.94837868e-01
2.13099197e-01 -5.46591401e-01 4.64468598e-01 -4.93779838e-01
5.68816543e-01 3.91892582e-01 -1.30896711e+00 9.69290078e-01
7.52854645e-01 3.18715483e-01 -5.03858387e-01 2.87118524e-01
6.31703734e-01 7.04955533e-02 -6.92125499e-01 3.47124308e-01
6.29098415e-01 -1.12182224e+00 6.83143377e-01 -2.45410159e-01
2.55319864e-01 -3.54883403e-01 3.53488535e-01 -1.09916413e+00
-1.56995654e-01 -9.61413145e-01 1.51992112e-01 7.98563123e-01
-1.36738643e-02 -1.01951838e+00 6.13819718e-01 -2.36935243e-01
3.43121350e-01 -7.45379746e-01 -1.41537964e+00 -1.12239945e+00
1.60337031e-01 7.56983161e-01 4.49203074e-01 5.58099985e-01
8.68239999e-02 9.12799835e-02 -3.14544104e-02 -2.72822350e-01
6.79158628e-01 7.74613842e-02 6.80196345e-01 -1.40562105e+00
-5.39291441e-01 -3.34227562e-01 -4.58036959e-02 -9.55572724e-01
-6.29663169e-01 -1.98334023e-01 1.42117754e-01 -1.67341936e+00
4.98375714e-01 -5.89123011e-01 6.95165023e-02 -2.52814204e-01
5.35815544e-02 5.88748008e-02 6.02269880e-02 3.63639414e-01
-1.15013309e-01 -3.07811704e-02 1.48110735e+00 5.03402412e-01
-5.30884206e-01 3.47823828e-01 6.23777583e-02 7.30804324e-01
7.74814963e-01 -6.85754180e-01 8.25655684e-02 5.11692464e-01
6.47491038e-01 7.34649301e-01 -4.94831018e-02 -7.48312354e-01
-2.59316862e-01 -3.90946567e-01 -2.70261526e-01 -7.61778593e-01
3.90488774e-01 -4.69860733e-01 1.00146019e+00 1.50789142e+00
6.21514499e-01 -1.22508042e-01 -3.80092531e-01 5.64712405e-01
6.41997084e-02 -8.62823129e-01 9.68861103e-01 -2.51598537e-01
4.81781840e-01 6.52146265e-02 -1.59087563e+00 -2.08240584e-01
1.42638934e+00 -8.10538232e-01 -2.77620494e-01 -1.18498728e-01
-9.52952325e-01 -1.38353482e-02 1.42391443e+00 -3.29924792e-01
-8.54502916e-02 -8.05740416e-01 -5.72823048e-01 -9.63814557e-02
-4.50948954e-01 1.32185910e-02 -2.14127768e-02 1.19658685e+00
-1.16014838e+00 4.46822166e-01 -4.90854830e-01 -7.19281971e-01
-1.17782438e+00 5.97783625e-01 2.37323835e-01 -7.23927855e-01
-4.68729995e-02 8.74687672e-01 5.31159520e-01 8.02720860e-02
-3.59718949e-01 -1.51089683e-01 -5.33963978e-01 2.45566219e-01
9.45460945e-02 7.80956149e-01 -4.62231338e-01 -6.82819426e-01
-3.34377378e-01 4.95288730e-01 4.36668307e-01 -2.99069822e-01
1.20608652e+00 -1.98895633e-01 -9.27353442e-01 6.38560772e-01
6.45798385e-01 2.55734473e-01 -9.27505255e-01 4.73709166e-01
-3.34712178e-01 -3.97731990e-01 -6.51362002e-01 -1.20395124e-01
-5.23004234e-01 4.24385548e-01 -3.58891264e-02 5.80894470e-01
8.77419710e-01 9.40814987e-02 6.16235852e-01 2.34471545e-01
7.22522080e-01 -9.26529050e-01 1.80302620e-01 7.09411263e-01
4.53524560e-01 -3.77735853e-01 -1.80237785e-01 -6.10766888e-01
-1.22499108e-01 7.37130404e-01 1.43320680e-01 -2.89259762e-01
4.94626641e-01 7.09261715e-01 -5.02604246e-01 -8.33634734e-02
-1.20269716e+00 2.01241940e-01 -7.33243346e-01 8.57562721e-01
3.39222103e-01 2.06057325e-01 -1.16778469e+00 3.29552144e-01
-1.33942336e-01 1.43942818e-01 1.02157640e+00 1.03713846e+00
-9.20385122e-01 -1.35306430e+00 -8.85375679e-01 2.42074132e-01
-6.58917665e-01 1.36431962e-01 -4.64067459e-01 8.69806647e-01
4.39953178e-01 7.18548596e-01 5.35775304e-01 5.89732409e-01
-3.36416155e-01 -1.50811067e-02 7.49674201e-01 -5.63065112e-01
-3.06084454e-01 3.08464736e-01 -5.52461576e-03 4.64080945e-02
-2.17685983e-01 -1.20851648e+00 -1.56428337e+00 -3.95541221e-01
-2.82342404e-01 3.24819446e-01 7.61778116e-01 6.47612393e-01
-1.66379511e-01 1.93233848e-01 6.59717560e-01 -3.25698555e-01
-3.50936413e-01 -6.05148375e-01 -1.01949036e+00 -1.97371855e-01
2.69500345e-01 -7.41236269e-01 -6.92761362e-01 4.94129211e-01] | [5.710201263427734, 4.243622779846191] |
215d2154-6808-42e9-b7ac-7b5e487743a9 | the-effects-of-just-in-time-delivery-on | 2212.12285 | null | https://arxiv.org/abs/2212.12285v1 | https://arxiv.org/pdf/2212.12285v1.pdf | The Effects of Just-in-time Delivery on Social Engagement: A Cluster Analysis | Fooji Inc. is a social media engagement platform that has created a proprietary "Just-in-time" delivery network to provide prizes to social media marketing campaign participants in real-time. In this paper, we prove the efficacy of the "Just-in-time" delivery network through a cluster analysis that extracts and presents the underlying drivers of campaign engagement. We utilize a machine learning methodology with a principal component analysis to organize Fooji campaigns across these principal components. The arrangement of data across the principal component space allows us to expose underlying trends using a $K$-means clustering technique. The most important of these trends is the demonstration of how the "Just-in-time" delivery network improves social media engagement. | ['Nathan Klarer', 'Raziel Ruíz', 'Moisés Ramírez'] | 2022-12-23 | null | null | null | null | ['marketing'] | ['miscellaneous'] | [-1.72091439e-01 1.72778189e-01 -7.13690460e-01 -4.16220397e-01
-3.80875140e-01 -6.26976430e-01 8.54099631e-01 3.84414345e-01
-3.89730811e-01 2.33007625e-01 5.61818659e-01 -8.99807096e-01
-6.27973437e-01 -1.12328291e+00 -4.49621201e-01 -1.83585331e-01
-3.40899110e-01 3.78930688e-01 -4.78245050e-01 -5.32938898e-01
5.22811770e-01 1.54180840e-01 -1.15378439e+00 3.39001864e-01
3.15670282e-01 5.17512202e-01 -2.16636658e-01 4.61063772e-01
-4.35723901e-01 5.97233772e-01 -4.41315114e-01 -4.58553135e-01
3.73365939e-01 -2.06178188e-01 -2.73094714e-01 -1.36048540e-01
1.33420631e-01 -9.24751237e-02 -2.22953945e-01 6.82985842e-01
1.49183318e-01 -7.76969492e-02 3.02899957e-01 -1.28141284e+00
-9.50956270e-02 8.85688543e-01 -6.49248958e-01 1.81517720e-01
4.00407970e-01 -1.07678153e-01 1.17926967e+00 -3.73739362e-01
1.01292598e+00 1.19338918e+00 5.60323596e-01 -1.50998697e-01
-1.58969283e+00 -1.10276437e+00 1.70886591e-01 -4.23961759e-01
-1.08748817e+00 -1.43107578e-01 1.16423213e+00 -8.96043837e-01
6.58129990e-01 7.37949431e-01 1.10898113e+00 6.25994086e-01
1.97313994e-01 7.18446076e-01 1.72033346e+00 -4.25559670e-01
1.96148559e-01 5.24226069e-01 6.76343143e-01 1.30437061e-01
1.94463700e-01 2.16597289e-01 -5.75658202e-01 -5.11249006e-01
7.12710619e-01 4.01122272e-01 6.55085385e-01 5.53441383e-02
-9.94063795e-01 1.52657676e+00 3.28052759e-01 3.80021930e-01
-3.70208532e-01 2.80887276e-01 2.11160824e-01 4.01898116e-01
9.16920900e-01 8.78489375e-01 2.05734074e-02 -3.89724314e-01
-8.80223811e-01 4.73775506e-01 6.49810076e-01 4.21324492e-01
8.33689928e-01 -4.06648844e-01 1.73547007e-02 5.28717160e-01
6.07886553e-01 2.63394445e-01 -1.84407998e-02 -8.32181811e-01
1.57458782e-01 1.00795305e+00 1.59296125e-01 -1.72920215e+00
-6.07497752e-01 -4.83559430e-01 4.01705503e-02 2.79916555e-01
3.89576018e-01 -6.01375401e-01 -2.23576412e-01 1.44748807e+00
2.70341694e-01 -9.76297483e-02 -1.67199999e-01 5.74729025e-01
6.17205620e-01 2.42660984e-01 3.40440184e-01 -8.54467824e-02
1.40118408e+00 -5.76231182e-01 -8.21219146e-01 4.00835909e-02
6.61632955e-01 -1.08828270e+00 1.08839369e+00 9.59427729e-02
-7.51995564e-01 -3.97712022e-01 -6.75763667e-01 7.75299907e-01
-4.75020379e-01 -2.54284918e-01 1.42200613e+00 9.63372171e-01
-7.87246287e-01 5.21627426e-01 -5.77391148e-01 -5.11822641e-01
3.49460810e-01 2.00884283e-01 -2.76430119e-02 3.18147123e-01
-8.41834128e-01 4.08082068e-01 -1.68566436e-01 -3.05852056e-01
-5.35732985e-01 -6.62697256e-01 -2.51075119e-01 -2.65823305e-01
2.45111763e-01 -1.66537493e-01 9.00543392e-01 -1.03159821e+00
-1.17821145e+00 5.66018403e-01 -3.26832943e-02 -8.90397951e-02
5.69181204e-01 4.23376746e-02 -7.68169940e-01 -4.68008108e-02
4.06777412e-01 3.68807137e-01 5.04444420e-01 -1.05705667e+00
-7.75419831e-01 -2.78603166e-01 2.61971176e-01 -5.92783988e-02
-6.63210869e-01 4.69449341e-01 1.70308977e-01 -3.01714092e-01
2.58098572e-01 -1.23838913e+00 -5.95685005e-01 -9.21297193e-01
-4.40546423e-01 -4.52474773e-01 7.98773110e-01 -4.04594690e-01
1.48151040e+00 -2.30798340e+00 -6.11584425e-01 7.31077015e-01
5.74458480e-01 -4.57967371e-01 1.13099426e-01 9.24448729e-01
-9.88007337e-02 5.07746637e-01 8.21203530e-01 -1.97267488e-01
8.17096233e-02 -6.64071217e-02 -8.41103941e-02 1.81945145e-01
-2.78381556e-02 7.84929514e-01 -9.37542796e-01 -2.46761858e-01
1.16995052e-01 1.64126635e-01 -6.59069061e-01 -2.60553539e-01
-4.40439023e-02 3.17354023e-01 -5.25926054e-01 4.37771946e-01
4.39402848e-01 -1.84199259e-01 5.54210663e-01 1.84909888e-02
-8.25961769e-01 1.66127861e-01 -7.95675695e-01 1.35348344e+00
3.44141126e-01 1.06754446e+00 -1.12075917e-01 -6.84319377e-01
1.09240246e+00 1.15636049e-03 9.88819897e-01 -7.55663991e-01
4.29363012e-01 9.18154493e-02 9.84078050e-02 -3.73286933e-01
8.95889819e-01 -1.64188892e-01 -5.11389732e-01 1.02112770e+00
-4.81133014e-01 3.20466369e-01 1.74177364e-01 7.14900076e-01
1.07375276e+00 -5.69330990e-01 -3.27286363e-01 -7.60755002e-01
-4.64725465e-01 6.68078244e-01 2.93945581e-01 6.68071449e-01
-3.54982764e-01 -2.04859555e-01 8.13708007e-01 -3.83461088e-01
-1.11016107e+00 -7.47507274e-01 -1.07785515e-01 1.21450388e+00
-2.32212245e-01 -7.11487532e-01 -3.06395859e-01 -3.85791689e-01
3.34010422e-01 7.15117514e-01 -8.64786267e-01 4.18026209e-01
-9.06388164e-02 -8.00731182e-01 2.39200741e-01 7.02995136e-02
3.14473331e-01 -5.56189299e-01 -4.35901821e-01 3.35476249e-01
-3.52613591e-02 -5.71089804e-01 -4.02410217e-02 -1.65917054e-01
-7.15508163e-01 -1.08536267e+00 -1.43468315e-02 -4.92660880e-01
5.70548236e-01 6.87379003e-01 8.25169086e-01 -3.12644690e-01
-2.69511372e-01 5.10996222e-01 -4.20574427e-01 -7.40752637e-01
-2.45432854e-01 -5.45035116e-02 -8.51561725e-02 4.34172377e-02
9.91490304e-01 -6.89990580e-01 -7.94815660e-01 4.49616611e-01
-5.78422189e-01 3.93715483e-04 1.60312653e-01 3.50239649e-02
1.93971410e-01 3.78477931e-01 7.55160570e-01 -9.05561626e-01
1.11837006e+00 -7.78050780e-01 -4.58015859e-01 -4.31603044e-01
-7.83784151e-01 -6.80668056e-01 -3.64338100e-01 -4.37672764e-01
-5.78911185e-01 -2.82499671e-01 2.98646301e-01 1.38316989e-01
1.74755648e-01 9.98649061e-01 4.13367301e-01 5.40039539e-02
7.48449147e-01 -5.32509804e-01 5.60911775e-01 -4.59070742e-01
3.93330723e-01 7.11276889e-01 -2.68455327e-01 -2.96554834e-01
6.07515514e-01 8.39652359e-01 -2.57723808e-01 -8.52707922e-01
-5.95286369e-01 -8.22215140e-01 -8.08693096e-02 -8.53201926e-01
7.89245725e-01 -9.67527866e-01 -1.40866709e+00 1.47551686e-01
-4.49046195e-01 -4.42166388e-01 -4.06560898e-01 7.90224791e-01
-2.24692523e-01 -2.14773938e-01 -5.20065725e-01 -9.27025616e-01
1.96478114e-01 -6.59100533e-01 1.69360995e-01 3.83418351e-01
-6.47285521e-01 -9.97405171e-01 5.19169509e-01 6.32988334e-01
7.39160180e-01 8.39913189e-01 4.40730721e-01 -7.79633343e-01
-2.65212506e-01 -4.60991561e-01 -3.52890074e-01 7.26545230e-02
-5.14953472e-02 2.16067657e-01 -5.87008953e-01 -2.97901779e-01
-1.54597148e-01 2.56076336e-01 3.49336654e-01 6.84179842e-01
4.48163956e-01 -2.82287329e-01 -5.30588388e-01 -9.00783837e-02
1.42396688e+00 3.51476640e-01 6.11782432e-01 7.02980042e-01
3.23549062e-01 1.01218975e+00 6.21931255e-01 5.27105570e-01
5.15969574e-01 3.97015780e-01 2.81133503e-01 -5.33918619e-01
2.53603011e-01 -4.18087453e-01 1.38768747e-01 6.25341177e-01
-1.98494956e-01 5.79296410e-01 -8.37848961e-01 5.49973071e-01
-1.91619992e+00 -1.22108948e+00 -5.56756139e-01 1.87133479e+00
3.06266516e-01 3.51837933e-01 7.23490238e-01 -1.02800839e-01
5.56753159e-01 1.57251224e-01 9.04073343e-02 -6.44926250e-01
-1.33449277e-02 8.13029185e-02 7.03991234e-01 4.45643276e-01
-7.51351953e-01 8.16059053e-01 7.44674301e+00 6.09866440e-01
-1.03293741e+00 3.25955629e-01 7.32055426e-01 -3.98735255e-01
-4.82983261e-01 4.08067226e-01 -7.59878457e-01 3.54675561e-01
1.02142751e+00 -3.31794947e-01 1.22778833e-01 9.29761827e-01
7.31959820e-01 -1.90689012e-01 -5.57199597e-01 3.20865810e-01
-2.53350049e-01 -1.77365088e+00 -6.73896253e-01 7.69905567e-01
9.15216565e-01 4.24444228e-02 1.85044765e-01 4.05947894e-01
7.64636278e-01 -7.70802975e-01 5.09329855e-01 4.46185231e-01
5.39603293e-01 -9.48453784e-01 1.50631994e-01 6.18375987e-02
-9.21488523e-01 -2.69367963e-01 -9.81763750e-03 -1.00101793e+00
1.24965899e-01 8.95035505e-01 -1.17365348e+00 2.00092360e-01
5.75984955e-01 5.19307494e-01 -3.97696078e-01 5.15163302e-01
5.02767861e-02 9.99933004e-01 -3.00418049e-01 -4.12962496e-01
5.03355324e-01 -5.14320970e-01 6.07360959e-01 1.15910184e+00
1.82912961e-01 1.51129425e-01 3.57847571e-01 5.48912287e-01
3.85163963e-01 2.43995979e-01 -6.12713814e-01 -7.64309645e-01
4.07913357e-01 1.38267648e+00 -1.08973920e+00 -1.93034858e-01
-1.57227337e-01 2.17295550e-02 -1.77555874e-01 3.12914282e-01
-4.64269996e-01 -3.38450521e-01 7.32011497e-01 8.87867749e-01
-1.22329801e-01 -6.00643516e-01 -4.60861504e-01 -5.89984059e-01
-5.72407365e-01 -6.68673217e-01 1.61745921e-01 -2.09322527e-01
-1.32287002e+00 1.37191877e-01 1.36822671e-01 -7.56232798e-01
2.17637882e-01 -2.59098321e-01 -7.71608114e-01 6.77349329e-01
-1.08897960e+00 -1.21608305e+00 -2.72247106e-01 6.89242005e-01
9.89066735e-02 -2.33304575e-01 7.75252402e-01 4.26394016e-01
-6.24798775e-01 3.17474633e-01 3.68093789e-01 1.15375929e-01
5.06503463e-01 -8.51304233e-01 5.79223074e-02 2.00646490e-01
-2.21360296e-01 8.24179947e-01 8.68226349e-01 -9.80247438e-01
-1.34401500e+00 -7.29575515e-01 1.19643116e+00 -4.90676790e-01
1.24268973e+00 -4.80119020e-01 -4.60229293e-02 8.45732629e-01
3.79325360e-01 -9.54498112e-01 1.57347858e+00 1.10823238e+00
-2.24458531e-01 -3.57928760e-02 -8.71855855e-01 8.26779842e-01
5.84210396e-01 -6.04480743e-01 -2.14361310e-01 7.23310113e-01
7.63730168e-01 4.18063492e-01 -1.12312913e+00 -1.47750184e-01
7.39480436e-01 -8.83073866e-01 7.72577763e-01 -8.07384610e-01
6.92396462e-01 1.96231473e-02 -2.47458383e-01 -1.00776994e+00
-3.60832334e-01 -1.04732859e+00 8.04601490e-01 1.21348751e+00
5.92673361e-01 -9.43333745e-01 1.26721048e+00 9.43254709e-01
-1.54389292e-02 -4.57426041e-01 -6.21130526e-01 -4.58834976e-01
-1.52488112e-01 -7.25936711e-01 4.76385325e-01 1.66662407e+00
5.82263589e-01 4.18940723e-01 -3.10307443e-01 -4.88624424e-01
6.41075909e-01 9.51634347e-02 1.20902872e+00 -1.39605534e+00
-4.24809128e-01 -5.44095457e-01 -2.98985600e-01 -5.57439685e-01
-4.20372784e-01 -7.59912491e-01 -5.46821535e-01 -1.24085164e+00
4.29215282e-01 -1.11537600e+00 -9.24051702e-02 1.58307984e-01
1.82365611e-01 1.51506513e-01 4.34974819e-01 5.64219654e-01
-3.25601578e-01 -1.92136049e-01 1.07314098e+00 2.11073216e-02
-7.80446589e-01 4.75993901e-02 -1.37041271e+00 4.94415283e-01
1.08824110e+00 -5.17567754e-01 -1.93811104e-01 1.66726530e-01
9.18075681e-01 -1.44340098e-01 2.80133456e-01 -5.62592447e-01
-3.26006450e-02 -3.21787596e-01 -6.96022809e-02 -5.84201157e-01
2.65247673e-01 -6.69368565e-01 7.03890324e-01 4.51280326e-01
-4.00864899e-01 -4.98329774e-02 2.08553314e-01 5.99500597e-01
3.54763903e-02 1.55387208e-01 5.42872511e-02 6.26820549e-02
-1.63656339e-01 -3.46901357e-01 -5.74418247e-01 -6.20346069e-01
1.17019391e+00 -2.08123073e-01 -7.32874751e-01 -5.56121528e-01
-1.09023654e+00 8.42117742e-02 1.93985030e-01 1.35652348e-01
-3.06731351e-02 -1.46875060e+00 -7.41275251e-01 7.70995691e-02
-1.37397215e-01 -7.89973199e-01 3.86749178e-01 1.00869203e+00
-2.80201346e-01 3.67406309e-01 -4.42527860e-01 -6.25647724e-01
-1.33431828e+00 3.96938652e-01 -3.61580461e-01 -5.40491402e-01
-2.29899824e-01 6.14315689e-01 -3.06854576e-01 -3.00275415e-01
-1.59181312e-01 4.22392428e-01 -3.53083044e-01 6.42061532e-01
5.27894855e-01 5.72537065e-01 -5.66504717e-01 -4.05832916e-01
-1.87413961e-01 1.00073189e-01 -4.14693616e-02 -5.36664069e-01
1.33916819e+00 -9.15612802e-02 -1.15524240e-01 5.45578301e-01
1.19475985e+00 3.31865728e-01 -8.53323042e-01 -1.09394677e-01
1.07858153e-02 -9.62866664e-01 1.38000891e-01 -4.81326818e-01
-7.42317498e-01 4.93588120e-01 6.35529876e-01 9.58581805e-01
5.07463157e-01 -2.52063945e-02 3.41084898e-01 -1.48024902e-01
2.78897421e-03 -1.51770389e+00 3.98230016e-01 3.04131415e-02
5.60321689e-01 -1.04677641e+00 8.25254098e-02 -4.34961826e-01
-6.59336507e-01 6.26273692e-01 -1.61938921e-01 -3.73097420e-01
1.16802871e+00 7.03308731e-02 1.48283064e-01 -1.02311456e+00
-3.07785779e-01 1.44290607e-02 -2.20317557e-01 3.31964105e-01
5.62262416e-01 9.28245604e-01 -1.07501757e+00 4.48206604e-01
-4.80315179e-01 -2.43194327e-01 4.88722324e-01 1.03637373e+00
-6.56232059e-01 -1.16458380e+00 -4.96295482e-01 6.80742621e-01
-3.36632788e-01 2.43371695e-01 -7.07564414e-01 1.07921350e+00
-7.47722611e-02 1.63788629e+00 3.95429850e-01 -8.47233891e-01
3.98561150e-01 -8.57873857e-02 -3.47941250e-01 -2.35389665e-01
-1.01098990e+00 8.39643776e-01 6.68294191e-01 -4.17240053e-01
-5.75042844e-01 -9.36063945e-01 -8.32797050e-01 -1.11115694e+00
-2.99731344e-01 5.18324673e-01 1.53732777e+00 3.91998172e-01
6.09564185e-01 6.73609674e-01 1.16140807e+00 -5.61757207e-01
-5.26698679e-02 -9.45204139e-01 -8.01558435e-01 4.89998221e-01
-4.41816688e-01 -4.50346470e-01 -3.45010132e-01 -4.15533662e-01] | [10.241959571838379, 6.679447650909424] |
bb8da3c2-1dc6-4a2d-b5e9-ad361ec532ed | 190513539 | 1905.13539 | null | https://arxiv.org/abs/1905.13539v4 | https://arxiv.org/pdf/1905.13539v4.pdf | Unsupervised Object Segmentation by Redrawing | Object segmentation is a crucial problem that is usually solved by using supervised learning approaches over very large datasets composed of both images and corresponding object masks. Since the masks have to be provided at pixel level, building such a dataset for any new domain can be very time-consuming. We present ReDO, a new model able to extract objects from images without any annotation in an unsupervised way. It relies on the idea that it should be possible to change the textures or colors of the objects without changing the overall distribution of the dataset. Following this assumption, our approach is based on an adversarial architecture where the generator is guided by an input sample: given an image, it extracts the object mask, then redraws a new object at the same location. The generator is controlled by a discriminator that ensures that the distribution of generated images is aligned to the original one. We experiment with this method on different datasets and demonstrate the good quality of extracted masks. | ['Thierry Artières', 'Mickaël Chen', 'Ludovic Denoyer'] | 2019-05-27 | unsupervised-object-segmentation-by-redrawing | http://papers.nips.cc/paper/9434-unsupervised-object-segmentation-by-redrawing | http://papers.nips.cc/paper/9434-unsupervised-object-segmentation-by-redrawing.pdf | neurips-2019-12 | ['unsupervised-object-segmentation'] | ['computer-vision'] | [ 8.21005762e-01 4.33059484e-01 1.85829401e-01 -2.89268643e-01
-4.81265128e-01 -1.02199316e+00 5.19936204e-01 -6.11848477e-03
-7.40437448e-01 7.27928460e-01 -6.40767217e-01 -2.78212354e-02
1.88387960e-01 -1.11249328e+00 -1.16082573e+00 -9.84384596e-01
2.75334567e-01 7.58581221e-01 8.00438225e-01 5.45985401e-02
2.01726541e-01 8.41317832e-01 -1.48359752e+00 3.28640789e-01
5.64980686e-01 6.44618273e-01 1.92052335e-01 8.76582503e-01
-8.34762529e-02 5.02092063e-01 -9.82990503e-01 -3.30630541e-01
7.51348555e-01 -5.94305634e-01 -8.31745386e-01 6.33986056e-01
4.09459710e-01 6.72893003e-02 7.19437152e-02 1.30832326e+00
1.08278312e-01 -5.17528097e-04 8.47330332e-01 -1.19328523e+00
-2.96016783e-01 6.08364940e-01 -4.27902490e-01 -1.41978532e-01
-1.01225428e-01 5.15611649e-01 6.75538838e-01 -3.18958610e-01
8.71914029e-01 7.86321342e-01 2.03531921e-01 7.16754079e-01
-1.70323050e+00 -3.81730497e-01 -2.96896640e-02 -1.77425727e-01
-1.18280768e+00 -1.34358406e-01 9.64825988e-01 -5.76145351e-01
2.69584004e-02 4.74957824e-01 5.34113407e-01 1.02564430e+00
-1.45655766e-01 6.00105703e-01 1.35024118e+00 -6.11819446e-01
5.40597379e-01 6.19359434e-01 -2.86653876e-01 3.20529968e-01
3.43767762e-01 5.42086102e-02 -1.14888668e-01 2.39166304e-01
6.39352143e-01 -8.69630128e-02 -2.86184609e-01 -8.14631045e-01
-9.84573185e-01 5.80428839e-01 6.14758611e-01 4.65820789e-01
-3.30850035e-01 2.76522338e-01 -3.14031728e-02 2.04837739e-01
1.57003328e-01 5.47810674e-01 -3.72272640e-01 4.14305210e-01
-1.20427465e+00 2.59547532e-01 7.23035693e-01 6.95802271e-01
1.04447055e+00 -1.97555780e-01 -1.31669924e-01 3.79504949e-01
2.33284421e-02 4.09268767e-01 3.94470841e-01 -7.22967207e-01
1.42385706e-01 8.01129162e-01 3.38508487e-01 -9.31068480e-01
-4.18222249e-02 -3.65586936e-01 -6.21240556e-01 9.54268038e-01
8.18095088e-01 -3.32045227e-01 -1.28979397e+00 1.70080566e+00
6.56352580e-01 -5.70033900e-02 8.72116722e-03 7.09559321e-01
3.79267722e-01 5.23079216e-01 -9.86914486e-02 1.46371946e-01
1.04881036e+00 -8.71382654e-01 -4.69915986e-01 -4.31197405e-01
2.22196415e-01 -6.77492261e-01 9.11233366e-01 5.53976357e-01
-1.00445235e+00 -6.29174292e-01 -1.18412673e+00 2.14126423e-01
-6.92609847e-01 2.14967981e-01 7.11619109e-02 9.02301550e-01
-8.33456814e-01 6.32846773e-01 -6.25930607e-01 -1.31907403e-01
6.57452345e-01 6.40045404e-01 -4.12001282e-01 1.86613813e-01
-6.74452245e-01 4.87203479e-01 9.97098565e-01 8.80950093e-02
-9.44684744e-01 -2.84886360e-01 -6.51676416e-01 6.93456158e-02
5.24997890e-01 -4.03739482e-01 8.51351917e-01 -1.83887172e+00
-1.48283124e+00 1.09653378e+00 2.97258794e-01 -6.54103875e-01
1.13564491e+00 2.05497906e-01 -8.18400458e-03 2.67124236e-01
-3.57768238e-02 9.19074416e-01 1.52518439e+00 -1.77067554e+00
-8.34456384e-01 -2.73982137e-01 -8.54038447e-02 -1.50418803e-01
-2.44710129e-02 -3.45756054e-01 -5.83197296e-01 -7.02198982e-01
-1.46726996e-01 -1.04684246e+00 -4.19133216e-01 1.61727369e-02
-7.18262434e-01 3.01416367e-01 1.02018154e+00 -5.23469388e-01
6.53355360e-01 -2.40457845e+00 1.35418400e-01 4.60646957e-01
5.39607890e-02 3.01639080e-01 -7.04620108e-02 6.89303800e-02
-1.91140905e-01 5.92555255e-02 -8.80432427e-01 -2.86708504e-01
-1.39373854e-01 2.36790761e-01 -2.68286109e-01 5.81450045e-01
3.99808884e-01 7.70220041e-01 -7.57161379e-01 -4.92064714e-01
3.33952159e-01 3.18890333e-01 -4.76051599e-01 3.34232509e-01
-7.27616549e-01 7.33518124e-01 -2.27212682e-01 -1.25437044e-02
9.04633999e-01 8.13447312e-02 2.35581920e-01 1.36793312e-02
1.38844430e-01 -4.47644204e-01 -1.45417857e+00 1.38406837e+00
-1.72845304e-01 5.36230803e-01 -1.60621759e-02 -1.10694444e+00
1.02696288e+00 1.32007658e-01 4.32849914e-01 -2.39788562e-01
2.35152140e-01 2.28734076e-01 1.96792230e-01 -3.59751195e-01
2.77439058e-01 -8.23076889e-02 -1.15172938e-01 5.52572012e-01
2.36946046e-02 -4.54709768e-01 3.26095968e-01 -2.54486389e-02
8.49485040e-01 1.98877886e-01 2.79875770e-02 -1.04271702e-01
5.85977256e-01 1.39537945e-01 1.95877641e-01 8.08210135e-01
1.80335164e-01 8.58859777e-01 6.99856818e-01 -3.38126391e-01
-1.28060460e+00 -9.05302942e-01 8.57783109e-02 4.96859610e-01
7.61003271e-02 3.89685512e-01 -1.28211355e+00 -1.19297361e+00
-7.73810819e-02 5.00008166e-01 -1.04567742e+00 -9.39198807e-02
-5.85844219e-01 -6.69214249e-01 4.28279698e-01 5.67012690e-02
6.11817062e-01 -1.45385242e+00 -9.88258660e-01 1.59418061e-01
2.35492811e-01 -1.20278931e+00 -3.43719006e-01 3.80572140e-01
-6.00193739e-01 -1.15219975e+00 -6.22700810e-01 -8.45270753e-01
1.24916351e+00 -2.73339242e-01 9.36442912e-01 6.91303015e-02
-4.45600897e-01 -9.02756900e-02 -3.70516747e-01 -4.22842205e-01
-7.80571282e-01 3.49087387e-01 -5.69935441e-01 8.61966133e-01
-3.97958197e-02 -5.14114082e-01 -4.55125570e-01 2.45194182e-01
-1.67099977e+00 6.00708835e-03 6.61075771e-01 5.68308830e-01
6.00763619e-01 5.69345415e-01 3.14885169e-01 -1.48935604e+00
7.24019483e-02 -7.23356307e-02 -9.58592594e-01 1.50471240e-01
-5.07483147e-02 3.10721546e-01 8.03914964e-01 -5.19252241e-01
-9.31828737e-01 8.32160830e-01 2.57610623e-02 -6.86641037e-02
-6.42529547e-01 -3.04804504e-01 -6.97880864e-01 -1.50618225e-01
6.92009032e-01 2.57953554e-01 -1.31360635e-01 -4.48697150e-01
6.59083009e-01 6.23340964e-01 7.84425080e-01 -2.97562689e-01
1.21393013e+00 8.45447183e-01 -1.77221149e-02 -6.27061486e-01
-5.80944657e-01 -1.12119325e-01 -1.09754956e+00 -1.40043423e-01
1.03204513e+00 -2.34371707e-01 -2.92690128e-01 5.96782982e-01
-9.83885348e-01 -7.38368869e-01 -7.52157807e-01 2.55525745e-02
-5.27532458e-01 4.04093936e-02 9.35165212e-02 -6.00190461e-01
-4.53897677e-02 -1.20203960e+00 7.82364428e-01 2.06075460e-01
3.88389975e-02 -9.13975894e-01 -3.61311436e-02 2.39014655e-01
4.82989997e-02 7.24447310e-01 8.40818226e-01 -8.45916271e-01
-8.48201573e-01 -3.87489825e-01 -2.41664518e-02 6.87422812e-01
3.19950402e-01 1.78159505e-01 -9.57830608e-01 -1.42155990e-01
1.21613830e-01 3.97127960e-03 7.85032690e-01 4.75435555e-02
1.32276523e+00 -3.85415763e-01 -3.71770680e-01 4.59179372e-01
1.72456801e+00 1.39201894e-01 7.39597023e-01 3.24479401e-01
7.17342019e-01 7.21769094e-01 2.43446931e-01 -2.63171587e-02
-2.43680969e-01 6.75329864e-01 6.50340199e-01 -3.97444367e-01
-1.68113008e-01 -1.16399437e-01 2.20550615e-02 -1.79500446e-01
2.04902485e-01 -3.80806178e-01 -6.29299641e-01 7.71093726e-01
-1.62263656e+00 -7.38609612e-01 -1.04210302e-01 2.35847521e+00
9.44697976e-01 4.75589007e-01 2.04117984e-01 3.74149501e-01
8.34192395e-01 6.67266771e-02 -5.60111701e-01 -3.34529042e-01
-1.01689070e-01 4.93395746e-01 8.69812071e-01 5.71115494e-01
-1.10853314e+00 1.01271331e+00 5.81134081e+00 7.56900668e-01
-1.35240293e+00 1.77177601e-02 6.70179665e-01 -3.34288850e-02
-1.74268514e-01 -2.78520212e-03 -5.12477875e-01 7.52614737e-01
3.89854074e-01 2.53744900e-01 4.39891189e-01 7.06257522e-01
-4.45571430e-02 -4.16177362e-01 -1.08311141e+00 5.98843932e-01
2.14938596e-02 -1.01178026e+00 8.35360065e-02 3.37753445e-02
1.01282167e+00 -3.40661585e-01 1.54948071e-01 -2.76141316e-01
4.10032928e-01 -1.03631926e+00 8.50821912e-01 4.32223976e-01
5.94311476e-01 -8.22627068e-01 5.20392537e-01 4.40292984e-01
-5.86143732e-01 1.18222743e-01 -1.52360916e-01 2.78105021e-01
3.24716456e-02 6.28980637e-01 -9.88310099e-01 2.97548413e-01
3.96566749e-01 2.20431909e-01 -8.15679073e-01 1.03848135e+00
-6.66673124e-01 5.12806296e-01 -3.44826102e-01 2.30717540e-01
2.53716797e-01 -2.29022205e-01 4.36120391e-01 1.18149841e+00
-9.93056893e-02 -5.03832817e-01 1.38310447e-01 1.02792132e+00
-1.93470120e-01 2.01417413e-02 -6.68107986e-01 1.25317127e-01
9.88745242e-02 1.37933731e+00 -1.35462642e+00 -3.95422518e-01
-8.86926129e-02 1.41141450e+00 1.58814833e-01 2.64287591e-01
-8.63924742e-01 -5.58074713e-01 7.93697163e-02 2.70867020e-01
8.57499838e-01 6.54813647e-02 -3.47567648e-01 -6.53042853e-01
2.20386788e-01 -9.72122729e-01 1.18326828e-01 -5.90496659e-01
-9.42110300e-01 7.57671893e-01 -1.90705910e-01 -1.04393697e+00
-1.37883142e-01 -5.57576597e-01 -3.73935163e-01 7.80191123e-01
-1.38802445e+00 -1.10352492e+00 -2.82092988e-01 5.91515481e-01
3.86553973e-01 5.18309772e-02 5.05004823e-01 1.52225554e-01
-4.32902485e-01 3.48336667e-01 1.36928812e-01 4.34586078e-01
4.94451642e-01 -1.53518164e+00 2.52269775e-01 1.09780741e+00
5.64858139e-01 7.46866912e-02 8.39696646e-01 -4.77869004e-01
-6.85668766e-01 -1.31246233e+00 3.51170450e-01 -5.75905025e-01
3.43454838e-01 -6.47693276e-01 -8.31780732e-01 7.33421206e-01
3.73297781e-01 2.75835603e-01 2.01788381e-01 -7.97107577e-01
5.22449836e-02 -2.23212913e-01 -1.39939630e+00 4.58376646e-01
4.65377480e-01 -1.64395809e-01 -3.43606234e-01 3.54651153e-01
4.87443864e-01 -3.69604170e-01 -3.34243059e-01 2.09743127e-01
1.66828781e-01 -1.06764424e+00 6.59603953e-01 -5.63962579e-01
3.81423622e-01 -6.88989460e-01 2.72309899e-01 -1.25346041e+00
2.46718958e-01 -5.69538116e-01 1.97056949e-01 1.32110322e+00
4.70985770e-01 -5.84082842e-01 9.51582193e-01 4.83310640e-01
4.14722234e-01 -3.43281597e-01 -6.98314428e-01 -7.33413458e-01
-1.32342055e-02 -1.83081940e-01 7.62320876e-01 6.78257823e-01
-8.08652878e-01 1.65352032e-01 -1.94207296e-01 5.02954960e-01
5.48358917e-01 2.50905335e-01 1.15381145e+00 -1.07424808e+00
-5.66984951e-01 -3.01515818e-01 -5.38409948e-01 -5.14077127e-01
1.19806603e-01 -7.20719397e-01 2.30886847e-01 -1.13122988e+00
-4.96216305e-02 -6.19158149e-01 -2.32651401e-02 3.88966411e-01
-2.10329071e-01 7.79955029e-01 1.40402362e-01 3.42727676e-02
-2.35952213e-01 2.16919947e-02 1.17145300e+00 -2.81268150e-01
-3.33756179e-01 3.22980642e-01 -4.24028486e-01 8.41334462e-01
8.89592171e-01 -7.90768623e-01 -3.41447055e-01 -3.27478051e-01
-1.78205408e-02 -4.72957700e-01 5.77242374e-01 -1.21162140e+00
2.66829580e-02 -8.41319934e-02 5.98278582e-01 -2.59155780e-01
1.04418151e-01 -1.36669981e+00 4.73866820e-01 5.27548969e-01
-2.92883813e-01 -3.47236425e-01 4.48058397e-02 3.90689224e-01
-5.81119582e-02 -7.29431331e-01 1.17779446e+00 -4.32484239e-01
-4.77604896e-01 1.27847463e-01 -2.11975798e-01 -3.69990282e-02
1.53169000e+00 -2.75205314e-01 2.29384854e-01 -2.77254041e-02
-8.76445293e-01 -1.32583678e-01 8.15017760e-01 8.45095143e-02
8.54000971e-02 -9.68291223e-01 -5.00071287e-01 3.87188226e-01
-1.77403400e-03 4.60681170e-01 2.63266489e-02 2.19794795e-01
-7.66397417e-01 -1.86808363e-01 -2.54159153e-01 -5.22595167e-01
-1.24806130e+00 1.03488910e+00 5.56923926e-01 -2.53364652e-01
-5.29100657e-01 5.60206175e-01 1.75442442e-01 -4.24824178e-01
6.51182607e-02 -2.53329128e-01 -1.23487391e-01 2.38474179e-02
3.10773641e-01 -1.34131148e-01 6.09707274e-02 -4.38271821e-01
9.16772783e-02 4.86717105e-01 6.79280236e-02 -3.76382172e-01
1.36823106e+00 1.43075511e-01 -1.62233666e-01 2.57680178e-01
1.07593668e+00 4.88986343e-01 -1.65038967e+00 -1.57864660e-01
1.32716149e-01 -6.38575196e-01 -3.24030846e-01 -7.95226753e-01
-1.48252738e+00 7.33978093e-01 6.80903792e-01 4.91313368e-01
1.15815604e+00 -7.23916367e-02 5.63799858e-01 -4.22548167e-02
2.72597522e-01 -1.00385940e+00 7.59107471e-02 -1.13960080e-01
5.09576440e-01 -1.07193160e+00 -1.26967266e-01 -4.98288065e-01
-4.67735648e-01 1.03491938e+00 4.59830791e-01 -4.58536714e-01
4.16168541e-01 3.88656884e-01 2.28973314e-01 -5.94607033e-02
-1.07330613e-01 -3.67658287e-01 2.46857673e-01 7.60396421e-01
-2.34236225e-01 3.77739370e-02 -1.38932496e-01 1.71184435e-01
-2.07927227e-01 -3.72610688e-02 6.37562752e-01 7.60457456e-01
-3.08732092e-01 -1.48786187e+00 -6.53907001e-01 5.60454838e-02
-5.58422923e-01 1.87855944e-01 -7.13884056e-01 8.46660256e-01
6.46142542e-01 5.24569988e-01 1.19311050e-01 3.67785208e-02
1.81111977e-01 7.69973695e-02 4.96442437e-01 -7.80967355e-01
-5.73310971e-01 -1.61822382e-02 -4.59499687e-01 -2.73109078e-01
-4.84124869e-01 -5.31467199e-01 -1.29342926e+00 3.19645017e-01
-1.24093637e-01 1.22855231e-01 7.80910492e-01 8.85942817e-01
-1.83991983e-03 7.17117071e-01 8.41243863e-01 -7.22404361e-01
-1.27004340e-01 -6.99188530e-01 -5.98583043e-01 7.93831408e-01
3.90414894e-01 -2.72331983e-01 -3.15350085e-01 7.37006068e-01] | [10.428079605102539, 0.028521932661533356] |
b722a599-16f6-46b1-b5bf-e0e9f9717798 | pifuhd-multi-level-pixel-aligned-implicit | 2004.00452 | null | https://arxiv.org/abs/2004.00452v1 | https://arxiv.org/pdf/2004.00452v1.pdf | PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization | Recent advances in image-based 3D human shape estimation have been driven by the significant improvement in representation power afforded by deep neural networks. Although current approaches have demonstrated the potential in real world settings, they still fail to produce reconstructions with the level of detail often present in the input images. We argue that this limitation stems primarily form two conflicting requirements; accurate predictions require large context, but precise predictions require high resolution. Due to memory limitations in current hardware, previous approaches tend to take low resolution images as input to cover large spatial context, and produce less precise (or low resolution) 3D estimates as a result. We address this limitation by formulating a multi-level architecture that is end-to-end trainable. A coarse level observes the whole image at lower resolution and focuses on holistic reasoning. This provides context to an fine level which estimates highly detailed geometry by observing higher-resolution images. We demonstrate that our approach significantly outperforms existing state-of-the-art techniques on single image human shape reconstruction by fully leveraging 1k-resolution input images. | ['Jason Saragih', 'Shunsuke Saito', 'Tomas Simon', 'Hanbyul Joo'] | 2020-04-01 | pifuhd-multi-level-pixel-aligned-implicit-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Saito_PIFuHD_Multi-Level_Pixel-Aligned_Implicit_Function_for_High-Resolution_3D_Human_Digitization_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Saito_PIFuHD_Multi-Level_Pixel-Aligned_Implicit_Function_for_High-Resolution_3D_Human_Digitization_CVPR_2020_paper.pdf | cvpr-2020-6 | ['3d-object-reconstruction-from-a-single-image', '3d-human-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 2.80454785e-01 1.42185301e-01 1.55118689e-01 -4.51098502e-01
-9.43661690e-01 -3.60058725e-01 5.49609780e-01 7.43187219e-02
-2.23374128e-01 4.37137127e-01 4.66133118e-01 -4.87141013e-02
6.67143688e-02 -9.24464703e-01 -9.37310338e-01 -1.75534323e-01
2.02718720e-01 9.34496105e-01 2.70675719e-01 -2.04038769e-01
2.86836028e-01 7.59865224e-01 -1.52253711e+00 4.88061696e-01
3.63886893e-01 1.04533935e+00 1.22393146e-01 7.58111238e-01
-7.19740475e-03 7.62714148e-01 -1.73196092e-01 -1.39867514e-01
3.58975530e-01 -2.03260258e-01 -7.23056316e-01 1.34609431e-01
8.77821565e-01 -8.31267476e-01 -3.89555454e-01 6.91872418e-01
5.55528641e-01 4.52031344e-02 5.75794041e-01 -5.21669447e-01
-7.14128852e-01 8.29930529e-02 -6.25660002e-01 9.75717232e-02
5.93206346e-01 2.38249913e-01 7.30806708e-01 -1.16190350e+00
8.01833272e-01 1.40310383e+00 9.41733658e-01 4.36494738e-01
-1.60645998e+00 -3.41399848e-01 1.69351950e-01 -1.30128518e-01
-1.37390733e+00 -5.88404298e-01 8.06574166e-01 -3.92607808e-01
1.46115065e+00 7.88769871e-02 9.33809102e-01 9.76175964e-01
8.80588070e-02 5.57181358e-01 1.21797228e+00 -3.62802655e-01
2.50815582e-02 -1.64296225e-01 -4.65076298e-01 7.58433878e-01
2.82744884e-01 4.32380468e-01 -5.65937996e-01 -8.55428204e-02
1.62540710e+00 -1.12906449e-01 -2.98490137e-01 -5.44472098e-01
-1.17167425e+00 5.34192443e-01 7.61980712e-01 1.19327120e-01
-6.55387402e-01 3.76436919e-01 8.43294114e-02 3.21625546e-02
2.71638662e-01 5.09869993e-01 -3.60967398e-01 -6.75321296e-02
-1.21075094e+00 5.37683010e-01 5.29804647e-01 7.59245515e-01
5.96854508e-01 3.77951749e-02 1.31556004e-01 6.08576238e-01
2.11545497e-01 3.84948939e-01 8.35902840e-02 -1.36583984e+00
3.95290852e-01 4.78838265e-01 2.92995930e-01 -1.07053936e+00
-5.72118521e-01 -5.08060336e-01 -7.15227425e-01 7.36591280e-01
5.95636010e-01 4.43735570e-01 -9.44162548e-01 1.57058346e+00
3.37187082e-01 -2.09746853e-01 -2.44521156e-01 1.23820424e+00
6.41425312e-01 3.63113344e-01 3.34387943e-02 3.43017370e-01
1.21705246e+00 -8.39146137e-01 -1.76126570e-01 -4.16551650e-01
6.48014694e-02 -6.57921672e-01 1.13051188e+00 4.93238777e-01
-1.56149924e+00 -6.94152772e-01 -1.09141231e+00 -5.83305240e-01
-2.53661156e-01 4.83691096e-02 5.73003352e-01 3.14103961e-01
-1.26883328e+00 8.06675255e-01 -8.68932426e-01 -3.38707507e-01
7.24773169e-01 2.91512519e-01 -6.13051057e-01 -1.54071599e-01
-6.92567766e-01 1.13902426e+00 6.17045015e-02 -9.99449268e-02
-6.52378619e-01 -9.02454436e-01 -9.45425272e-01 8.51555765e-02
2.97221094e-01 -1.01322377e+00 1.40867400e+00 -7.86876142e-01
-1.28737295e+00 1.06958151e+00 -1.39496215e-02 -4.09182042e-01
7.21104562e-01 -2.50262648e-01 1.39621079e-01 3.87384504e-01
-6.41212538e-02 9.43083227e-01 7.69439697e-01 -1.62814987e+00
-3.25929523e-01 -5.90079248e-01 1.77526265e-01 3.11305225e-01
2.01390803e-01 -3.37374181e-01 -3.04090410e-01 -5.68293333e-01
4.12880242e-01 -6.47827923e-01 -3.40632886e-01 4.68577653e-01
7.08838552e-02 1.76084787e-01 5.86583853e-01 -6.21484578e-01
6.67152166e-01 -1.64778984e+00 5.31424433e-02 -5.62923867e-03
2.90980905e-01 7.26435930e-02 4.90303896e-03 3.04908425e-01
6.25623837e-02 -8.92615914e-02 -1.76109925e-01 -6.81818724e-01
-1.07727364e-01 1.40170559e-01 -5.06111205e-01 3.82496059e-01
2.47999832e-01 1.15267241e+00 -7.40802586e-01 -5.78352571e-01
5.19100130e-01 8.91664267e-01 -7.33497560e-01 2.14143053e-01
-3.11666518e-01 5.77840090e-01 -3.09702754e-01 6.47580147e-01
7.01740682e-01 -6.14527762e-01 1.70418441e-01 -5.31212091e-01
-1.05262183e-01 3.28867048e-01 -1.00215709e+00 2.17192268e+00
-4.45630372e-01 3.58910024e-01 1.63382560e-01 -8.96331906e-01
7.35215247e-01 1.83589756e-01 4.28510606e-01 -8.10724318e-01
6.17592633e-02 2.53700912e-01 -2.89498329e-01 -2.63558567e-01
6.89889371e-01 -3.58287126e-01 3.63877192e-02 5.42219698e-01
-2.14820847e-01 -5.52047372e-01 -5.63986778e-01 2.11028848e-02
9.15246189e-01 8.16460788e-01 5.38134515e-01 -3.25311385e-02
2.22551093e-01 2.04789504e-01 2.16621026e-01 6.60698354e-01
-2.43494481e-01 1.15687287e+00 1.15510620e-01 -9.34208930e-01
-1.67743576e+00 -1.22331655e+00 -3.62997651e-02 8.82364571e-01
1.42116651e-01 -2.76042819e-01 -5.47352612e-01 -2.66961575e-01
6.86550662e-02 5.16201854e-01 -7.95627892e-01 2.39492893e-01
-9.60985899e-01 -3.05715173e-01 4.96022433e-01 1.12936699e+00
5.27883470e-01 -1.08144295e+00 -1.25901461e+00 2.86641747e-01
4.06282097e-02 -1.38702846e+00 -1.93624049e-01 -1.29740730e-01
-1.09147048e+00 -8.32712173e-01 -7.75777161e-01 -4.39921230e-01
7.51619995e-01 2.03154132e-01 1.67416525e+00 1.80851713e-01
-5.22642672e-01 5.13362408e-01 8.71159788e-03 -1.97699536e-02
-3.27493250e-01 -1.47501871e-01 2.98855994e-02 -5.65599620e-01
2.95412391e-02 -8.84415984e-01 -8.69537950e-01 9.31713656e-02
-7.02969134e-01 4.12446380e-01 9.31774199e-01 8.10825646e-01
9.38988507e-01 -2.71798611e-01 3.77815008e-01 -6.77612305e-01
2.86809862e-01 -3.20330933e-02 -5.98070681e-01 1.46962404e-01
-5.77172101e-01 2.36630648e-01 6.26840889e-01 -2.47775480e-01
-1.12206066e+00 3.37595791e-01 -3.04803163e-01 -6.67950571e-01
-4.00087059e-01 2.13355988e-01 3.46665710e-01 -1.57539696e-01
7.87426651e-01 3.56425554e-01 2.81910934e-02 -4.39551532e-01
2.23377511e-01 2.66349047e-01 7.54640341e-01 -8.19423854e-01
5.74661732e-01 7.32387781e-01 2.70324588e-01 -6.44589961e-01
-9.81320381e-01 -1.82811227e-02 -9.67448294e-01 -1.52864292e-01
8.19166958e-01 -1.09520495e+00 -7.10615337e-01 1.54979423e-01
-1.05243623e+00 -4.62810040e-01 -3.32286090e-01 2.70556509e-01
-8.88536692e-01 2.18387753e-01 -6.63815200e-01 -9.05888796e-01
-4.88379955e-01 -1.05062401e+00 1.48936319e+00 1.36684086e-02
-4.73609209e-01 -6.27837479e-01 -1.42524928e-01 4.39624906e-01
5.95175266e-01 6.10786319e-01 8.32833648e-01 1.02216095e-01
-1.00265288e+00 -1.14574127e-01 -5.49831510e-01 -1.53772697e-01
-2.23259032e-01 -4.05691117e-01 -1.13794446e+00 -2.67831653e-01
-1.35848835e-01 -7.86106288e-01 6.67693377e-01 4.67632324e-01
1.28566253e+00 -1.10422634e-02 -2.34531119e-01 6.11089349e-01
1.58165288e+00 -4.48018819e-01 5.74153781e-01 1.93327114e-01
7.16227174e-01 5.43420672e-01 3.75406563e-01 2.54579812e-01
6.80553496e-01 7.87766099e-01 3.95114243e-01 -1.58472404e-01
-4.86878604e-01 -5.71866572e-01 -1.18523009e-01 3.68912965e-01
-4.39126909e-01 2.51156747e-01 -1.04340327e+00 4.64836270e-01
-1.79302597e+00 -1.02487516e+00 2.90940613e-01 2.28788519e+00
6.77197516e-01 3.16582590e-01 2.36554265e-01 6.46743849e-02
9.68797132e-02 9.72993374e-02 -8.57030988e-01 -2.73426324e-01
1.42011018e-02 2.41372332e-01 3.02772313e-01 6.17204905e-01
-7.11364388e-01 7.65843332e-01 7.15335894e+00 4.52063769e-01
-1.15197408e+00 -1.36240542e-01 7.71324039e-01 -3.18172485e-01
-3.36363316e-01 -2.43524313e-01 -4.28545773e-01 -8.62180889e-02
6.31658912e-01 2.87678421e-01 5.13080060e-01 1.00260365e+00
-6.19335158e-04 -2.27966845e-01 -1.27320075e+00 1.26691866e+00
1.31899685e-01 -1.49542975e+00 8.40942860e-02 2.61658043e-01
6.85437202e-01 1.39721960e-01 1.48051426e-01 9.70825106e-02
6.37542680e-02 -1.61328757e+00 9.99279201e-01 7.05279052e-01
1.29368174e+00 -6.58452034e-01 2.45138794e-01 5.24228394e-01
-1.35951674e+00 9.42426268e-03 -4.98452902e-01 -2.51243174e-01
1.85108215e-01 2.86889344e-01 -7.84006655e-01 2.70085394e-01
7.29034007e-01 3.35317880e-01 -3.52513760e-01 5.52813888e-01
1.62605569e-02 4.48469520e-02 -5.13904989e-01 2.78031290e-01
9.43027064e-02 1.74822614e-01 2.44009718e-01 1.08856332e+00
1.85014918e-01 5.17995119e-01 2.35992581e-01 1.16665649e+00
-5.52397743e-02 -2.70391047e-01 -6.91658318e-01 1.84419483e-01
4.83631521e-01 1.04156828e+00 -7.35201478e-01 -4.26294923e-01
-4.42509025e-01 1.08860898e+00 8.33086073e-01 1.33460179e-01
-5.95369279e-01 2.64975280e-01 5.71397841e-01 7.08846211e-01
5.13245106e-01 -6.54701591e-01 -8.28663886e-01 -1.04482031e+00
2.39582881e-02 -7.87525475e-01 1.88998073e-01 -1.03942776e+00
-1.12317550e+00 6.90442383e-01 -7.10176304e-02 -9.98523474e-01
-3.99325788e-01 -7.61547029e-01 -1.28642052e-01 9.73163426e-01
-1.52109420e+00 -1.44076669e+00 -4.96333480e-01 3.72422874e-01
6.06526673e-01 2.08963379e-01 1.15046430e+00 8.57603252e-02
1.72483355e-01 4.77043033e-01 -4.91488725e-01 9.04641226e-02
4.12616670e-01 -1.21974790e+00 6.77891374e-01 5.63905895e-01
1.56245470e-01 5.11340618e-01 6.59681380e-01 -6.26616120e-01
-1.43201756e+00 -5.35892546e-01 8.16667080e-01 -8.24663818e-01
-5.14283478e-02 -3.54190499e-01 -9.11766410e-01 6.12004638e-01
-2.02187657e-01 4.74736750e-01 2.18562096e-01 1.41464863e-02
-6.22550011e-01 2.90056795e-01 -1.43846297e+00 5.83384693e-01
1.27299881e+00 -7.12014019e-01 -6.72123909e-01 -2.12999076e-01
4.82597351e-01 -9.12998617e-01 -1.03340018e+00 4.32821840e-01
9.93677497e-01 -1.40870607e+00 1.44500387e+00 -4.25202012e-01
8.33186626e-01 -2.71552622e-01 -3.32246989e-01 -9.61525679e-01
-5.44416189e-01 -7.63122588e-02 -4.15023595e-01 3.86656016e-01
6.80118278e-02 -2.28900298e-01 1.04874742e+00 8.99303019e-01
1.33023620e-01 -1.17726958e+00 -8.07334840e-01 -4.47873473e-01
2.91473955e-01 -4.65878189e-01 7.57578790e-01 6.39601648e-01
-2.69385725e-01 2.31043160e-01 -2.80558884e-01 1.08897723e-01
9.02794242e-01 5.81967235e-01 6.97074890e-01 -1.18411243e+00
-3.62902194e-01 -5.36928117e-01 -3.60002667e-01 -1.38068199e+00
-1.97009861e-01 -4.75053728e-01 7.83184096e-02 -1.55267382e+00
1.03599742e-01 -4.49989021e-01 1.64230287e-01 3.35131556e-01
4.89756092e-02 7.70184398e-01 1.34876251e-01 3.18762153e-01
-3.74436349e-01 4.38814193e-01 1.43234229e+00 3.29218209e-01
1.39393777e-01 -4.69776899e-01 -6.90758228e-01 8.98928285e-01
5.32998443e-01 1.37231871e-02 -2.95198411e-01 -8.30424368e-01
2.95084119e-01 3.01304251e-01 8.00326347e-01 -1.16308165e+00
1.29587293e-01 -8.18858445e-02 1.29318666e+00 -8.23692083e-01
8.37102711e-01 -7.93083668e-01 1.51869640e-01 3.08265716e-01
-4.77748901e-01 7.56556764e-02 2.10788235e-01 4.47758436e-01
1.87351972e-01 2.99159110e-01 1.02321243e+00 -5.89812517e-01
-5.83156526e-01 4.40898865e-01 1.12036459e-01 -5.25606051e-02
5.42413592e-01 -5.46114028e-01 4.67611402e-02 -5.10337532e-01
-6.72786057e-01 -2.73791879e-01 9.72303033e-01 1.86627671e-01
9.45152640e-01 -1.35838652e+00 -6.75632894e-01 3.53442222e-01
-3.03449519e-02 2.99810737e-01 2.65591323e-01 5.41132390e-01
-5.83930850e-01 4.38386798e-01 -4.14136499e-01 -6.90371335e-01
-9.50965464e-01 4.35306787e-01 5.26442409e-01 -2.97057152e-01
-1.28670835e+00 7.99347520e-01 3.23124617e-01 -4.56493378e-01
6.04845546e-02 -3.37752044e-01 1.34711549e-01 -6.10281527e-01
6.13255441e-01 5.65440245e-02 4.62563448e-02 -8.89711738e-01
-2.72033960e-01 9.92697895e-01 1.30816311e-01 -2.27703735e-01
1.42384386e+00 -1.93837404e-01 4.03404534e-01 1.60364732e-01
9.84165251e-01 -9.46581587e-02 -1.74560940e+00 -2.45014682e-01
-4.01195914e-01 -7.36668646e-01 1.64747357e-01 -1.00397587e+00
-7.57341206e-01 8.86836946e-01 4.45677072e-01 -6.39178753e-02
1.02787280e+00 1.10800214e-01 1.07970941e+00 4.02443588e-01
6.83988631e-01 -9.81821060e-01 1.67863324e-01 3.89126986e-01
1.18563509e+00 -1.41526008e+00 3.84126037e-01 -3.31293553e-01
-4.30868626e-01 1.30394936e+00 7.22700059e-01 -4.89741415e-01
3.98490191e-01 4.09926444e-01 -1.61891300e-02 -5.11750579e-01
-6.16691887e-01 -9.91886333e-02 4.39538509e-01 6.80998743e-01
5.64119279e-01 8.43132474e-03 2.63367116e-01 3.76660705e-01
-5.03085852e-01 9.32414308e-02 3.44826609e-01 8.20951939e-01
-4.33723867e-01 -8.11250925e-01 -6.15703404e-01 3.43243301e-01
-2.16507599e-01 1.09852023e-01 -2.31492177e-01 6.35157883e-01
1.17887661e-01 4.43467468e-01 2.15288326e-01 -1.42375290e-01
4.11390424e-01 -2.14748189e-01 1.17186081e+00 -4.94803101e-01
-3.59896928e-01 -7.56049529e-02 -8.29421449e-04 -9.74897563e-01
-3.00868124e-01 -5.69791615e-01 -1.37720156e+00 -3.05040598e-01
2.50860780e-01 -4.81333762e-01 5.49678624e-01 8.46252441e-01
5.09348154e-01 4.16265607e-01 -3.03976219e-02 -1.57809472e+00
-6.02525234e-01 -8.16597760e-01 -1.83501929e-01 4.29784328e-01
4.54506427e-01 -8.26016247e-01 8.46764594e-02 3.48898955e-03] | [8.770164489746094, -3.285536766052246] |
b7bd9485-4c81-41b8-8006-f9b45dec9d4f | cross-lingual-linking-of-automatically | null | null | https://aclanthology.org/2022.lrec-1.713 | https://aclanthology.org/2022.lrec-1.713.pdf | Cross-lingual Linking of Automatically Constructed Frames and FrameNet | A semantic frame is a conceptual structure describing an event, relation, or object along with its participants. Several semantic frame resources have been manually elaborated, and there has been much interest in the possibility of applying semantic frames designed for a particular language to other languages, which has led to the development of cross-lingual frame knowledge. However, manually developing such cross-lingual lexical resources is labor-intensive. To support the development of such resources, this paper presents an attempt at automatic cross-lingual linking of automatically constructed frames and manually crafted frames. Specifically, we link automatically constructed example-based Japanese frames to English FrameNet by using cross-lingual word embeddings and a two-stage model that first extracts candidate FrameNet frames for each Japanese frame by taking only the frame-evoking words into account, then finds the best alignment of frames by also taking frame elements into account. Experiments using frame-annotated sentences in Japanese FrameNet indicate that our approach will facilitate the manual development of cross-lingual frame resources. | ['Ryohei Sasano'] | null | null | null | null | lrec-2022-6 | ['cross-lingual-word-embeddings'] | ['natural-language-processing'] | [ 1.22480266e-01 3.66861939e-01 -2.24989593e-01 -5.90876639e-01
-7.96518624e-01 -4.51751530e-01 8.05586100e-01 4.26063329e-01
-6.23894989e-01 7.56255805e-01 7.30289578e-01 -2.99447656e-01
1.24957912e-01 -7.97515750e-01 -3.84429038e-01 -2.65595704e-01
2.45140150e-01 3.12617451e-01 6.40911996e-01 -2.78671086e-01
1.07228629e-01 1.02454059e-01 -1.37909198e+00 3.25577855e-01
2.83640265e-01 5.62864602e-01 6.21292830e-01 2.26342306e-01
-7.12387085e-01 9.33741212e-01 -5.07396638e-01 -6.64325297e-01
-2.20220700e-01 -7.31680632e-01 -1.25621772e+00 3.39409083e-01
-1.41393989e-01 2.31193617e-01 7.57517517e-02 9.30017531e-01
1.81591496e-01 1.89872637e-01 8.39086324e-02 -9.46792722e-01
-7.00803876e-01 9.73735213e-01 1.29019171e-01 3.30307066e-01
5.70742846e-01 -4.78366196e-01 1.21766949e+00 -8.18066537e-01
1.06868732e+00 1.42098665e+00 4.23376560e-01 5.71649313e-01
-9.98916209e-01 -1.26482993e-01 8.79423693e-02 3.82317752e-01
-1.20257652e+00 -4.54277039e-01 8.78096342e-01 -4.59728390e-01
1.31156039e+00 -1.28119275e-01 7.56419718e-01 7.05315530e-01
1.85736213e-02 5.70836663e-01 6.60106361e-01 -1.23210216e+00
1.62866302e-02 6.11416288e-02 3.92636955e-01 5.15600562e-01
9.31316465e-02 -4.08969373e-01 -4.80473101e-01 2.59117372e-02
6.80499613e-01 -5.56606829e-01 -3.02781854e-02 -1.79619342e-01
-1.40287828e+00 7.22750366e-01 -8.41858331e-03 9.93994951e-01
-1.61898643e-01 6.02633879e-02 7.92315185e-01 -7.87891224e-02
6.11536920e-01 3.52850974e-01 -3.76770467e-01 -2.68997997e-01
-4.38279897e-01 3.17405999e-01 5.73293686e-01 9.18219864e-01
8.10124755e-01 7.25890463e-03 3.53629172e-01 8.08622122e-01
3.90799880e-01 1.90497667e-01 6.47200227e-01 -8.81975055e-01
4.49752480e-01 7.65889406e-01 3.45726669e-01 -1.03997743e+00
-2.70271868e-01 3.85027885e-01 2.03307778e-01 -4.07194465e-01
4.06519145e-01 8.94112629e-04 -3.95481408e-01 1.85875785e+00
5.85020304e-01 -6.33278415e-02 5.28763652e-01 5.68991184e-01
7.92409837e-01 5.68140268e-01 4.19710815e-01 -4.74438250e-01
1.72218919e+00 -6.46130800e-01 -1.04260337e+00 -1.07069321e-01
1.00608063e+00 -9.05791223e-01 1.01445532e+00 -3.43992859e-01
-1.09878469e+00 -6.41822755e-01 -8.35779130e-01 -3.82715374e-01
-5.73697150e-01 -8.78283903e-02 3.60067517e-01 3.90251011e-01
-6.42941773e-01 -6.75075594e-03 -1.03064466e+00 -6.23584747e-01
-1.34546906e-01 -3.04860622e-03 -4.63654369e-01 1.57622084e-01
-1.48959863e+00 1.26535881e+00 7.93293655e-01 7.18098581e-02
-3.45814645e-01 -1.56479806e-01 -1.37256074e+00 -1.80593699e-01
5.81737578e-01 -2.63588309e-01 1.21161032e+00 -1.28867900e+00
-1.15854323e+00 9.19447064e-01 -5.01535833e-01 -6.44521475e-01
-2.08204672e-01 -2.62063473e-01 -9.60268736e-01 1.45321772e-01
5.38333654e-01 5.74197650e-01 4.05661881e-01 -8.79600406e-01
-1.00376618e+00 -4.52654622e-02 6.62492335e-01 2.46214896e-01
-1.40345186e-01 7.85470963e-01 -3.98975104e-01 -7.48767436e-01
-2.79056262e-02 -6.71737373e-01 8.69012177e-02 -5.14513075e-01
1.62349045e-01 -7.21489251e-01 6.82062089e-01 -5.72812140e-01
1.52236915e+00 -2.18730783e+00 1.39037326e-01 -1.19718581e-01
-2.46938050e-01 -1.22402668e-01 8.89933109e-02 3.71869415e-01
-4.99616228e-02 8.22977200e-02 7.44090825e-02 -5.82655743e-02
-1.20969005e-01 7.44043469e-01 -2.56197661e-01 1.20220430e-01
1.98650986e-01 5.04935384e-01 -1.19906592e+00 -8.29066336e-01
3.94302517e-01 3.99267048e-01 -5.67728460e-01 -6.85956180e-02
-3.38877231e-01 4.14563417e-02 -5.85768700e-01 3.93959701e-01
-2.52073467e-01 6.52549695e-03 7.44324386e-01 -3.96572858e-01
-5.05321205e-01 7.57671237e-01 -1.13286150e+00 1.56961203e+00
-5.26577115e-01 7.26810753e-01 -5.48390448e-01 -9.40447450e-01
7.64397323e-01 9.72313523e-01 5.21171808e-01 -4.64884013e-01
2.43746459e-01 3.55699867e-01 1.61799863e-02 -5.85902214e-01
6.91024423e-01 -4.23899144e-01 -4.31547582e-01 3.52898955e-01
2.72619307e-01 -9.93265212e-02 7.02064395e-01 -1.24874145e-01
5.11198044e-01 6.25092864e-01 7.37042964e-01 -3.33562493e-01
8.16530228e-01 2.86417425e-01 7.89712608e-01 -1.23361237e-02
-7.67643005e-02 3.12767714e-01 3.79640073e-01 -9.38023508e-01
-9.84435737e-01 -4.23678428e-01 -1.12837873e-01 1.02246678e+00
2.52624869e-01 -1.03977323e+00 -9.86231148e-01 -7.06915975e-01
-5.79224825e-01 9.16088820e-01 -4.12222594e-01 3.90978843e-01
-1.04487050e+00 -3.86767447e-01 4.67606753e-01 5.66835821e-01
2.91337430e-01 -1.27540898e+00 -8.57559204e-01 7.32495964e-01
-4.34546918e-01 -1.32734537e+00 -2.88006663e-01 -1.01890430e-01
-2.61328697e-01 -1.34606910e+00 -1.71002001e-01 -9.71995056e-01
5.55360496e-01 1.26749784e-01 9.75988686e-01 -6.72552688e-03
2.44659454e-01 2.54878461e-01 -6.57895148e-01 -5.35100996e-01
-5.19986987e-01 -7.39619583e-02 1.16911523e-01 -5.27401939e-02
7.64560699e-01 1.20782264e-01 1.88859865e-01 4.68821675e-01
-1.00195289e+00 1.90752283e-01 -4.30005372e-01 5.73596001e-01
6.25975311e-01 7.33721405e-02 4.82818604e-01 -7.69747496e-01
4.38426822e-01 -4.74982075e-02 -6.90126836e-01 5.44383466e-01
5.26375957e-02 1.22139893e-01 1.25537112e-01 -2.32021704e-01
-1.39596951e+00 -4.69659939e-02 -1.41846284e-01 1.47603050e-01
1.64272394e-02 7.50531852e-01 -4.20315951e-01 4.47110116e-01
5.84774673e-01 -1.13473371e-01 -3.22324693e-01 -2.29467407e-01
5.05222678e-01 3.21212173e-01 3.77194583e-01 -9.33728755e-01
4.69630897e-01 1.36506721e-01 -3.59571785e-01 -1.07974827e+00
-1.02265275e+00 -1.47607997e-01 -8.50391388e-01 -3.41325015e-01
1.39902556e+00 -9.34837639e-01 9.89077613e-02 2.68153585e-02
-1.39134514e+00 -1.27086803e-01 -4.35951889e-01 6.92141116e-01
-5.35769343e-01 3.42349321e-01 -2.79628754e-01 -3.80238324e-01
1.73735142e-01 -1.26770234e+00 7.40028143e-01 3.87519412e-02
-8.78521979e-01 -1.36153996e+00 1.09529890e-01 2.31229633e-01
-1.27003610e-01 1.28344730e-01 1.02980983e+00 -7.37197340e-01
-1.87028483e-01 5.70328608e-02 1.43993482e-01 2.26484612e-01
5.93647003e-01 8.96054953e-02 -5.18071592e-01 1.23731367e-01
-3.42968805e-03 -1.61085144e-01 -7.00713173e-02 2.31583774e-01
3.66546810e-01 -2.97282618e-02 -3.13418120e-01 -1.14415996e-01
1.35885632e+00 6.94211602e-01 3.98533970e-01 5.85093915e-01
6.13364458e-01 5.97005844e-01 6.70879066e-01 -4.48889285e-02
9.10040438e-01 9.31079924e-01 -3.33257854e-01 2.24370107e-01
-1.17866442e-01 -1.44451380e-01 3.41493487e-01 1.09675848e+00
-2.06141800e-01 -2.72736430e-01 -1.02277017e+00 9.00092125e-01
-1.88508606e+00 -1.10467291e+00 -9.27608386e-02 1.60361814e+00
1.05212665e+00 3.99348110e-01 8.37888867e-02 8.01537484e-02
8.99866998e-01 1.98618338e-01 2.01563463e-01 -2.36493349e-01
-1.12020403e-01 1.17146000e-01 5.67855276e-02 8.54727387e-01
-1.05302000e+00 1.45103216e+00 6.18720055e+00 5.21542132e-01
-8.81657958e-01 3.21436375e-01 9.07977298e-02 3.82949203e-01
-3.97834778e-01 4.39879060e-01 -9.80064332e-01 3.55671555e-01
9.48136866e-01 -4.91260499e-01 4.18116190e-02 7.87873626e-01
3.80705684e-01 -1.49682179e-01 -1.09860027e+00 7.20054030e-01
1.73070028e-01 -1.75292909e+00 8.22776482e-02 -3.64243686e-01
4.79216099e-01 -2.36351267e-01 -8.54412317e-01 4.67219986e-02
2.83571720e-01 -1.93562806e-01 1.26354718e+00 1.35879099e-01
5.43097198e-01 -6.77366495e-01 8.21150124e-01 -2.13394538e-01
-1.56480789e+00 4.71016437e-01 -2.92338848e-01 -2.03218207e-01
7.14753389e-01 2.28800982e-01 -6.29995763e-01 5.37633955e-01
3.28633457e-01 6.14395261e-01 -3.47269416e-01 4.18252498e-01
-3.92539024e-01 6.06261492e-01 -2.32315317e-01 -4.73090559e-02
2.91181505e-01 -3.48597988e-02 3.62005830e-01 1.19632661e+00
3.00587624e-01 2.51264453e-01 5.37876070e-01 4.40510303e-01
2.43165106e-01 3.10485065e-01 -5.83594978e-01 -9.22853798e-02
5.46717167e-01 7.95736611e-01 -1.13346970e+00 -6.76988363e-01
-9.42839861e-01 4.05194938e-01 8.38629305e-02 6.54310435e-02
-8.43733072e-01 -4.52114433e-01 6.68493569e-01 1.59343481e-01
3.19744297e-03 -5.42867661e-01 4.32728492e-02 -1.29412258e+00
-8.41342583e-02 -7.01202989e-01 4.12092149e-01 -9.30316210e-01
-7.71306336e-01 8.93732727e-01 5.21322310e-01 -9.39531326e-01
-5.49237192e-01 -4.01848197e-01 -2.29646504e-01 7.22320676e-01
-9.46729779e-01 -1.28541696e+00 2.15517968e-01 4.32579935e-01
1.05884814e+00 -6.58547580e-02 9.46134150e-01 2.58332878e-01
-4.04037178e-01 3.12765241e-02 -5.61169922e-01 3.77923161e-01
6.08978093e-01 -9.20550585e-01 3.96176040e-01 1.11921358e+00
4.18297172e-01 6.68436110e-01 8.70877266e-01 -7.52844512e-01
-7.85768926e-01 -8.43165338e-01 1.88294399e+00 -4.97953624e-01
1.08746421e+00 -1.54991701e-01 -9.12726283e-01 1.01494348e+00
4.57158089e-01 -6.85421526e-02 9.06811953e-01 5.30332886e-02
-7.06288964e-02 9.11465031e-04 -7.48431742e-01 6.93413734e-01
9.62436378e-01 -7.98147440e-01 -1.39313948e+00 1.43482611e-01
8.40561509e-01 -4.37066406e-01 -7.89760888e-01 1.29245505e-01
2.64424026e-01 -4.15509969e-01 6.97284818e-01 -4.25014436e-01
1.14007592e-01 -7.85850525e-01 -4.01246369e-01 -9.09609497e-01
-5.90515211e-02 -4.26886559e-01 3.80896240e-01 1.46755612e+00
4.00722712e-01 -2.68361002e-01 2.45206863e-01 7.59628952e-01
-4.00965989e-01 -4.65989485e-02 -9.85132515e-01 -5.08518100e-01
-3.34473252e-01 -8.24096441e-01 5.78085661e-01 1.13672054e+00
4.72034395e-01 6.32054925e-01 -6.39538690e-02 7.87562951e-02
7.68108815e-02 -7.91018754e-02 4.62328732e-01 -1.14017880e+00
1.16509281e-01 -8.89475867e-02 -4.67886776e-01 -4.64892596e-01
6.37769401e-01 -8.43222380e-01 5.13917282e-02 -1.52557242e+00
-3.75938416e-01 -3.53463650e-01 -2.07149573e-02 7.62489676e-01
-7.20920041e-02 3.57053503e-02 1.98688060e-01 1.61970168e-01
-5.33289373e-01 2.51729369e-01 8.01859498e-01 2.26736546e-01
-2.02552557e-01 -5.25740504e-01 -4.21806276e-01 1.09168756e+00
5.47880590e-01 -5.33622503e-01 -5.61378539e-01 -6.49384856e-01
2.84501374e-01 1.07854553e-01 -3.72750089e-02 -9.05785620e-01
-2.61293054e-02 -4.20807332e-01 -5.95832691e-02 -2.75442719e-01
1.27176180e-01 -9.95961607e-01 5.32515228e-01 1.48256958e-01
-3.45905244e-01 4.94398117e-01 1.61984384e-01 1.63402230e-01
-5.80152452e-01 -5.90959609e-01 5.25545657e-01 -3.39211524e-01
-1.12076378e+00 -4.52938437e-01 -6.63560569e-01 2.10197479e-01
1.12110198e+00 -2.44283542e-01 -7.51403645e-02 1.27240896e-01
-9.16559577e-01 -1.20791130e-01 4.43301499e-01 5.78978300e-01
2.78226644e-01 -1.54399753e+00 -1.27045333e-01 1.18576333e-01
2.01980919e-01 -8.12015682e-02 -2.26204842e-01 3.05633187e-01
-6.17081285e-01 4.02854919e-01 -2.19666854e-01 -2.35192403e-01
-1.27121592e+00 5.26250720e-01 7.72336647e-02 -4.85070273e-02
-6.04726136e-01 4.03599679e-01 -1.18790865e-01 1.08333817e-02
-1.11876763e-01 -6.98274553e-01 -4.33133900e-01 3.50372910e-01
4.61428434e-01 8.29950124e-02 -1.88239917e-01 -1.32661116e+00
-4.76759702e-01 5.73484659e-01 3.45842987e-01 -3.69550854e-01
1.19346428e+00 -3.80576074e-01 -3.02803189e-01 7.34425545e-01
8.16512644e-01 1.43383026e-01 -7.98285544e-01 -4.12703872e-01
6.38128161e-01 -2.51527399e-01 -1.58949211e-01 -6.78301156e-02
-5.24971128e-01 3.49954009e-01 1.22410998e-01 4.28161770e-01
8.12150300e-01 2.04332098e-01 7.25113809e-01 2.61339873e-01
7.09161103e-01 -1.27439380e+00 -1.06476106e-01 7.49338031e-01
6.32881999e-01 -7.92203188e-01 -1.58075124e-01 -5.53848982e-01
-5.20762682e-01 1.21282363e+00 5.13799667e-01 1.77760087e-02
7.96189845e-01 6.90031871e-02 1.55757830e-01 -1.96855888e-01
-5.47452927e-01 -4.57976639e-01 1.12451702e-01 3.87804687e-01
8.07863295e-01 6.33271262e-02 -9.29425895e-01 3.94776851e-01
-1.58985585e-01 2.92308480e-01 5.36072969e-01 1.22024047e+00
-5.92004001e-01 -1.61961699e+00 -2.65993088e-01 -2.90723026e-01
-6.68242812e-01 -8.03406611e-02 -1.54300317e-01 1.12873852e+00
5.68631768e-01 8.27935815e-01 4.73582715e-01 5.54664955e-02
3.16039592e-01 4.81078893e-01 4.68423039e-01 -9.63045180e-01
-4.10696089e-01 4.53634471e-01 6.77739978e-01 -2.76256531e-01
-1.38666475e+00 -7.21110463e-01 -1.45799756e+00 1.26335055e-01
-3.92751902e-01 4.55977380e-01 3.39662701e-01 1.47490525e+00
3.49197909e-02 3.74871701e-01 6.44216239e-02 -7.50024796e-01
4.81946737e-01 -7.22992957e-01 -3.71673182e-02 4.34293747e-01
-1.14124231e-01 -8.61565650e-01 3.36992368e-02 7.50438035e-01] | [10.23812484741211, 9.296518325805664] |
b6fb260a-384c-46e6-85bd-8898c26de6a1 | discriminative-invariant-kernel-features-a | null | null | http://openaccess.thecvf.com/content_cvpr_2016/html/Pal_Discriminative_Invariant_Kernel_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Pal_Discriminative_Invariant_Kernel_CVPR_2016_paper.pdf | Discriminative Invariant Kernel Features: A Bells-and-Whistles-Free Approach to Unsupervised Face Recognition and Pose Estimation | We propose an explicitly discriminative and `simple' approach to generate invariance to nuisance transformations modeled as unitary. In practice, the approach works well to handle non-unitary transformations as well. Our theoretical results extend the reach of a recent theory of invariance to discriminative and kernelized features based on unitary kernels. As a special case, a single common framework can be used to generate subject-specific pose-invariant features for face recognition and vice-versa for pose estimation. We show that our main proposed method (DIKF) can perform well under very challenging large-scale semi-synthetic face matching and pose estimation protocols with unaligned faces using no land-marking whatsoever. We additionally benchmark on CMU MPIE and outperform previous work in almost all cases on off-angle face matching while we are on par with the previous state-of-the-art on the LFW unsupervised and image-restricted protocols, without any low-level image descriptors other than raw-pixels. | ['Felix Juefei-Xu', 'Dipan K. Pal', 'Marios Savvides'] | 2016-06-01 | null | null | null | cvpr-2016-6 | ['unsupervised-face-recognition'] | ['computer-vision'] | [ 5.09778738e-01 3.81626077e-02 1.24934845e-01 -6.69160306e-01
-1.01867819e+00 -6.03064060e-01 7.78706968e-01 -5.24289608e-01
-3.83758008e-01 5.50802469e-01 1.15466081e-02 4.02105004e-02
-3.92298341e-01 -7.86761403e-01 -8.97834539e-01 -9.38522577e-01
-6.13438226e-02 8.01057577e-01 4.36321944e-02 -3.57358277e-01
2.00729095e-03 1.18018138e+00 -1.98478091e+00 -4.66986932e-02
2.52619058e-01 8.72530997e-01 -3.30336273e-01 6.30647361e-01
5.47200859e-01 1.21556260e-01 -1.61429599e-01 -6.08273268e-01
6.64606035e-01 -2.08251938e-01 -7.01226473e-01 6.66038096e-02
1.47694612e+00 -5.89628555e-02 -4.35220212e-01 8.74393463e-01
1.06168818e+00 -1.97904054e-02 7.90426016e-01 -1.07993269e+00
-4.18822855e-01 4.15765941e-02 -4.16853875e-01 -1.22522488e-01
6.13916218e-01 4.84592170e-02 5.60628295e-01 -9.98117745e-01
1.03303516e+00 1.49411762e+00 1.11924207e+00 7.35522509e-01
-1.59652436e+00 -7.95993745e-01 -4.06802148e-01 1.80989474e-01
-1.50285649e+00 -1.11073041e+00 5.56922138e-01 -2.23981127e-01
8.22016120e-01 5.40224969e-01 1.88174024e-01 1.05471098e+00
-2.74176542e-02 2.33594239e-01 1.50769770e+00 -6.91948175e-01
1.59436092e-02 -2.13050067e-01 -2.91819274e-01 7.75626242e-01
3.29250246e-01 4.37542915e-01 -5.61086416e-01 -3.32762718e-01
8.68382812e-01 -3.24127436e-01 -1.64380491e-01 -1.00980711e+00
-1.24801016e+00 6.02882326e-01 2.88735896e-01 1.29246488e-01
-8.08625445e-02 1.59220904e-01 2.76011348e-01 6.68931663e-01
4.60166931e-01 5.57756536e-02 -5.66301703e-01 1.35479465e-01
-1.25969934e+00 5.14851093e-01 7.83258975e-01 1.06471586e+00
1.11877871e+00 1.50694311e-01 -3.23157907e-01 6.15825593e-01
2.83913732e-01 1.06443024e+00 -5.26028238e-02 -8.97870839e-01
9.68025401e-02 4.33662981e-02 1.98561475e-02 -6.95416272e-01
-4.11100119e-01 -3.35303873e-01 -6.18275940e-01 4.50667679e-01
5.36924303e-01 4.85147871e-02 -9.63157892e-01 1.87952375e+00
5.79477251e-01 3.20701152e-01 -1.21359795e-01 5.15006006e-01
6.82127774e-01 -5.57290651e-02 -1.57679811e-01 -2.58126080e-01
1.32828033e+00 -5.87888658e-01 -4.53893483e-01 -4.32761349e-02
4.41526175e-01 -1.21893978e+00 5.72907686e-01 1.54592186e-01
-8.04741919e-01 -4.93295908e-01 -8.39712739e-01 2.95984417e-01
-2.98737437e-01 2.96488792e-01 9.08934653e-01 1.29647672e+00
-1.45757997e+00 9.41216290e-01 -5.72386444e-01 -8.11572433e-01
4.39723134e-01 7.30842113e-01 -1.06274152e+00 -2.01852888e-01
-9.02141750e-01 8.81092012e-01 1.37630075e-01 1.67981476e-01
-7.69578874e-01 -9.21963871e-01 -1.05017328e+00 -4.72940087e-01
2.53854871e-01 -5.65864325e-01 1.05941129e+00 -9.02637780e-01
-1.63625181e+00 1.41721010e+00 -9.17795002e-02 -2.42056698e-02
8.04981172e-01 4.34348397e-02 -3.84700775e-01 1.36866033e-01
-1.06013559e-01 8.66970122e-01 1.14932299e+00 -9.92693007e-01
1.52295560e-01 -7.04393089e-01 -1.06668495e-01 -8.13373029e-02
-2.25315094e-01 4.10604715e-01 -3.63733113e-01 -6.74454510e-01
1.10392431e-02 -1.14377010e+00 1.44771859e-01 4.22413170e-01
-3.95670384e-02 -8.56618136e-02 1.01575625e+00 -6.35851204e-01
6.15849376e-01 -1.72242475e+00 2.37235263e-01 4.12576437e-01
-4.47242886e-01 3.05527747e-01 -5.06659687e-01 4.31898832e-01
-5.48806548e-01 -4.85593706e-01 -3.53319466e-01 -5.08696198e-01
4.45703357e-01 1.79721355e-01 -1.21967748e-01 1.21863723e+00
7.28848055e-02 9.78593111e-01 -6.09684050e-01 -4.62824553e-01
4.13388819e-01 7.77848840e-01 -7.51910627e-01 1.39831811e-01
2.13122800e-01 5.03579736e-01 -6.78961724e-02 7.55629897e-01
1.38697147e+00 5.52964747e-01 9.35788155e-02 -5.71040869e-01
-7.70724416e-02 -2.23535478e-01 -1.46849430e+00 1.78959942e+00
-5.93463540e-01 4.86830622e-01 3.94541711e-01 -1.11722565e+00
7.99309671e-01 3.28015596e-01 4.78672862e-01 -3.99126977e-01
-3.93156707e-02 3.91557097e-01 -2.04122558e-01 -8.10084790e-02
-1.87086597e-01 -1.46444708e-01 1.71004921e-01 2.88675815e-01
6.43087745e-01 -2.24553794e-01 -6.10952005e-02 -2.17332035e-01
1.01034451e+00 7.58250654e-01 3.30295146e-01 -8.93817842e-01
8.27800274e-01 -5.31888723e-01 1.43333599e-01 5.93506396e-01
-1.68873593e-01 8.08875859e-01 1.71014488e-01 -4.10417140e-01
-1.01509917e+00 -1.21182692e+00 -7.96107054e-01 1.12554216e+00
-4.04891968e-01 -2.20399648e-01 -1.19369459e+00 -6.03729606e-01
2.38657191e-01 1.55804768e-01 -8.59625399e-01 9.70793217e-02
-5.29170275e-01 -8.74297619e-01 7.98739016e-01 2.61679560e-01
4.54308242e-01 -1.09411716e+00 -1.59623668e-01 -1.69404119e-01
4.42226648e-01 -1.11772382e+00 -4.58772302e-01 1.25431251e-02
-4.23343807e-01 -1.13545966e+00 -8.43379080e-01 -7.10882783e-01
8.12158644e-01 5.98913953e-02 1.06887889e+00 -2.06289381e-01
-7.49918222e-01 9.38802004e-01 -5.18020391e-02 -1.93511069e-01
-1.97539732e-01 -8.95928442e-02 4.61196661e-01 2.81224400e-01
2.93128192e-01 -6.39273524e-01 -7.01469600e-01 6.00975692e-01
-8.63171816e-01 -5.11327446e-01 5.47751844e-01 7.21619129e-01
3.51995319e-01 -7.00473487e-01 2.55947798e-01 -9.12631512e-01
1.44901248e-02 1.47871166e-01 -7.14458704e-01 4.52294528e-01
-3.48129243e-01 2.27463245e-01 4.72730070e-01 -3.20199728e-01
-1.12409627e+00 5.44416785e-01 -3.69981080e-01 -2.58677214e-01
-4.54319686e-01 -4.47787553e-01 -5.43287039e-01 -1.25328875e+00
8.28751981e-01 2.16271043e-01 6.43519610e-02 -4.02541339e-01
4.66070950e-01 6.08179271e-01 6.26621008e-01 -8.79655004e-01
1.53830290e+00 8.64719570e-01 6.95524573e-01 -1.15644443e+00
-3.28602672e-01 -4.88559693e-01 -1.18600643e+00 -2.17257917e-01
5.88085115e-01 -8.79675150e-01 -8.84185433e-01 6.95669174e-01
-9.76160288e-01 -1.42362341e-01 -2.24843144e-01 3.43907058e-01
-9.26266313e-01 6.64935470e-01 -3.56220007e-01 -6.00510538e-01
-3.12076479e-01 -1.07410085e+00 1.78589475e+00 8.32943339e-03
1.28824085e-01 -9.80734050e-01 3.09875220e-01 2.49181390e-01
6.34107232e-01 3.46204787e-01 2.60446459e-01 -2.20391229e-01
-4.32928771e-01 -3.35130960e-01 -2.33938158e-01 2.56193846e-01
-8.45429152e-02 -5.92808202e-02 -1.46123266e+00 -8.55684996e-01
-2.51251459e-01 -5.46664596e-01 8.30933154e-01 5.24499305e-02
8.04147899e-01 4.09415737e-02 -2.80736208e-01 1.08051729e+00
1.30944645e+00 -5.35608053e-01 8.96532297e-01 1.40912961e-02
5.57229221e-01 7.50225842e-01 4.73293573e-01 3.16810340e-01
-1.26943722e-01 1.35818815e+00 2.20069841e-01 -2.21314430e-01
-2.95114607e-01 1.09293386e-02 6.95671916e-01 3.61189336e-01
-5.79530239e-01 4.52832192e-01 -5.95519781e-01 1.27931252e-01
-1.54308975e+00 -1.10700941e+00 2.54096482e-02 2.51506257e+00
4.97307420e-01 -5.49450636e-01 -2.74020154e-02 -2.33604554e-02
5.54880977e-01 1.96592599e-01 -9.91826504e-02 -2.31357440e-01
-1.70144573e-01 1.00467503e+00 8.56093764e-01 5.09251654e-01
-1.51475751e+00 1.03351068e+00 6.92011356e+00 1.30574799e+00
-1.04745328e+00 3.69880587e-01 2.16196969e-01 2.76751757e-01
-2.57240474e-01 5.69825508e-02 -9.72192287e-01 -8.65908638e-02
7.98984826e-01 2.85731137e-01 5.26478708e-01 8.82859111e-01
-2.97174096e-01 7.74692520e-02 -1.11121380e+00 1.11019909e+00
6.33066952e-01 -1.01481140e+00 -7.46447667e-02 1.19341411e-01
7.73425996e-01 -1.22620672e-01 1.24811880e-01 1.38009563e-01
-1.17063932e-01 -1.06126177e+00 4.63571548e-01 4.34633017e-01
1.35117221e+00 -7.08796680e-01 4.01377380e-01 -2.38627195e-02
-1.51264238e+00 4.35373008e-01 -7.57659912e-01 2.40641207e-01
-2.25359231e-01 2.33181611e-01 -4.51981425e-01 8.18967760e-01
3.26513410e-01 4.41330791e-01 -8.86015654e-01 8.20273280e-01
-1.53776035e-01 2.84352303e-01 -6.73599422e-01 4.77048010e-01
-1.69019159e-02 -2.98799396e-01 4.65416610e-01 1.38441479e+00
4.97175336e-01 -3.71119916e-01 -1.11620463e-01 3.45544547e-01
-4.83658304e-03 3.32390100e-01 -8.12797487e-01 5.13131201e-01
1.66786723e-02 1.59908605e+00 -7.28577375e-01 1.15284532e-01
-3.84979755e-01 1.23465574e+00 3.13688695e-01 2.46879384e-01
-7.13158131e-01 -1.72425881e-01 7.70773470e-01 2.19432577e-01
4.47166562e-01 -1.48053706e-01 5.68184435e-01 -1.25603974e+00
1.07483603e-01 -6.45157397e-01 2.15140298e-01 -4.08781439e-01
-1.09239578e+00 6.22761726e-01 3.59403759e-01 -1.10000157e+00
-4.21175301e-01 -1.18993235e+00 -5.39078295e-01 8.34764183e-01
-1.60257053e+00 -2.02810884e+00 -1.83728904e-01 7.73288190e-01
2.37783045e-02 -3.03191721e-01 1.39615595e+00 5.91909826e-01
-1.62286952e-01 1.03871143e+00 2.75983125e-01 -1.20321205e-02
1.20002306e+00 -1.02977586e+00 5.14359295e-01 8.93704832e-01
4.42726582e-01 6.49602652e-01 6.82410181e-01 -4.02797341e-01
-1.87947047e+00 -1.07583082e+00 6.76541269e-01 -8.23322237e-01
4.38333690e-01 -9.36838865e-01 -4.29420561e-01 7.74689913e-01
1.55422734e-02 7.33231604e-01 4.19745445e-01 9.56799686e-02
-6.34132743e-01 -4.10545260e-01 -1.39333677e+00 3.72259438e-01
1.62789190e+00 -6.61881089e-01 -3.29760522e-01 7.10581839e-01
-4.79967473e-03 -3.98431122e-01 -8.96824956e-01 7.09089518e-01
8.84583294e-01 -1.05816734e+00 1.43581009e+00 -4.89256829e-01
-3.92943740e-01 -2.26772055e-01 -2.60390639e-01 -9.85404789e-01
-3.06875587e-01 -1.05168629e+00 3.04199636e-01 1.20160091e+00
-3.26064117e-02 -7.24213362e-01 9.02427435e-01 1.99055329e-01
1.66154474e-01 -2.86183864e-01 -1.53492117e+00 -1.05053556e+00
6.85163662e-02 -3.71017098e-01 5.15755236e-01 7.34483123e-01
-5.43368638e-01 -1.64792001e-01 -5.51480234e-01 1.19336016e-01
1.08165205e+00 1.14365838e-01 9.80031967e-01 -1.27186000e+00
-1.34709731e-01 -1.45366713e-01 -9.92805302e-01 -5.78288555e-01
7.54240394e-01 -1.03890193e+00 -2.13611126e-01 -6.92957938e-01
2.61282325e-01 -1.49638012e-01 7.79638737e-02 4.64295208e-01
2.21212387e-01 9.21357572e-01 -1.45303737e-02 -1.67466834e-01
-4.43912685e-01 3.09475392e-01 1.12026119e+00 9.73419845e-02
5.10871649e-01 -3.69884521e-02 -1.19346552e-01 5.22554755e-01
4.28780675e-01 -2.44942680e-01 -1.83481276e-01 1.89784933e-02
1.43107949e-02 -4.30516928e-01 5.27525246e-01 -1.31010258e+00
1.15604341e-01 1.88305024e-02 5.02283692e-01 -1.15494229e-01
3.87610823e-01 -9.07016516e-01 6.48057222e-01 3.73829424e-01
2.34771073e-01 -1.90073162e-01 2.56467015e-01 2.40154341e-01
7.65136927e-02 -1.11761421e-01 1.14234555e+00 1.00105554e-01
-7.20565975e-01 6.91834867e-01 5.36641777e-02 -1.91051006e-01
1.15986359e+00 -3.08292001e-01 -1.19330078e-01 -3.31964105e-01
-5.81384599e-01 -4.52070773e-01 8.60861540e-01 2.85816789e-01
3.27430338e-01 -1.66852176e+00 -9.69795942e-01 7.43733823e-01
4.33511436e-01 -6.52625442e-01 5.29097319e-01 9.04337704e-01
-9.16055858e-01 6.42712116e-01 -5.24404585e-01 -5.31157136e-01
-1.59289551e+00 4.58395571e-01 4.48288918e-01 -1.05685234e-01
-2.42177621e-01 6.80154085e-01 6.67389929e-02 -1.09114873e+00
-2.73202769e-02 1.28945500e-01 5.41177541e-02 2.04444751e-01
4.77567434e-01 1.71061814e-01 5.67258835e-01 -1.46109653e+00
-6.32345021e-01 1.29121673e+00 2.90968180e-01 4.87985387e-02
1.35304749e+00 3.39942545e-01 -2.13811308e-01 -2.75257438e-01
1.38146770e+00 2.15956345e-01 -1.15292525e+00 -1.45364106e-01
-3.03167611e-01 -6.02313280e-01 -2.03901991e-01 -3.13815564e-01
-1.07826531e+00 8.11480701e-01 1.03988647e+00 -4.93121415e-01
8.91613185e-01 -2.16540713e-02 1.04953386e-01 6.51426077e-01
7.26719975e-01 -9.35010076e-01 -2.75643796e-01 2.89962202e-01
1.12839854e+00 -1.25837910e+00 2.22503603e-01 -8.09364140e-01
-7.66803995e-02 1.33101583e+00 4.23980564e-01 -1.86100632e-01
7.96286345e-01 4.04370487e-01 3.17402780e-02 -1.30482269e-02
-1.08760871e-01 -4.48092222e-01 6.24088228e-01 1.07360888e+00
4.96593893e-01 4.84636463e-02 -8.21508765e-02 -2.89233506e-01
-1.27825752e-01 -1.89792901e-01 -1.70415536e-01 8.64600062e-01
-1.06685661e-01 -1.59885716e+00 -7.28402495e-01 1.74096629e-01
-2.72299886e-01 4.53451127e-02 -2.91222334e-01 9.73472059e-01
2.06247896e-01 2.72914171e-01 -3.76394354e-02 -1.18850537e-01
4.48380738e-01 2.36231148e-01 1.32433093e+00 -5.45639753e-01
-4.48098749e-01 -1.03206523e-01 -7.16592371e-02 -9.14946198e-01
-8.27483416e-01 -9.96746957e-01 -4.87295449e-01 -4.17044610e-01
-1.39843553e-01 -2.29423568e-01 4.76777077e-01 6.97882056e-01
1.45628840e-01 -3.95564847e-02 6.07930541e-01 -1.48025823e+00
-7.75140047e-01 -1.02587163e+00 -6.74949646e-01 8.07850063e-01
8.53129029e-02 -9.99980748e-01 -3.65623593e-01 -1.26460686e-01] | [13.215163230895996, 0.2831770181655884] |
308f33ee-fa2b-4c00-ace5-35f78355abf0 | bi-directional-differentiable-input | 1811.01116 | null | http://arxiv.org/abs/1811.01116v2 | http://arxiv.org/pdf/1811.01116v2.pdf | Bi-Directional Differentiable Input Reconstruction for Low-Resource Neural Machine Translation | We aim to better exploit the limited amounts of parallel text available in
low-resource settings by introducing a differentiable reconstruction loss for
neural machine translation (NMT). This loss compares original inputs to
reconstructed inputs, obtained by back-translating translation hypotheses into
the input language. We leverage differentiable sampling and bi-directional NMT
to train models end-to-end, without introducing additional parameters. This
approach achieves small but consistent BLEU improvements on four language pairs
in both translation directions, and outperforms an alternative differentiable
reconstruction strategy based on hidden states. | ['Xing Niu', 'Weijia Xu', 'Marine Carpuat'] | 2018-11-02 | bi-directional-differentiable-input-1 | https://aclanthology.org/N19-1043 | https://aclanthology.org/N19-1043.pdf | naacl-2019-6 | ['low-resource-neural-machine-translation'] | ['natural-language-processing'] | [ 4.84854192e-01 1.86823756e-01 -7.55515277e-01 -3.32343996e-01
-1.56675887e+00 -7.30871081e-01 1.03508294e+00 -3.09961438e-01
-5.42083204e-01 1.09719741e+00 4.43532974e-01 -7.88519323e-01
4.51061487e-01 -4.75514054e-01 -1.27892327e+00 -7.66841844e-02
5.07475317e-01 9.36524689e-01 -4.22967643e-01 -6.77414089e-02
-3.77838984e-02 1.76265329e-01 -7.16078877e-01 5.45137107e-01
8.82630169e-01 4.31953877e-01 2.64081240e-01 6.05523467e-01
-8.51658285e-02 4.38738793e-01 -3.20070595e-01 -8.11173260e-01
4.46428716e-01 -9.01608527e-01 -9.31778431e-01 -1.89993396e-01
3.69902015e-01 -2.59803563e-01 -3.16832364e-01 7.77540505e-01
5.04852235e-01 -1.84442475e-02 5.56235433e-01 -6.16909206e-01
-1.02224600e+00 7.29853749e-01 -1.31143853e-01 1.46440491e-01
3.72772545e-01 3.40027988e-01 1.12746155e+00 -1.23418462e+00
9.04944897e-01 1.31074333e+00 6.03892922e-01 6.89457953e-01
-1.61490571e+00 -3.50737125e-01 -2.12973431e-01 -1.70857787e-01
-8.63772869e-01 -8.04549456e-01 3.50485831e-01 -1.61836401e-01
1.44862270e+00 2.29345933e-01 9.92505625e-02 1.66166365e+00
4.16423678e-01 8.71041954e-01 1.27648032e+00 -6.80237591e-01
-1.31114200e-01 1.08998500e-01 -3.50073725e-01 5.99075675e-01
-1.49401322e-01 2.31577903e-01 -7.22644985e-01 -2.12353662e-01
6.47505760e-01 -1.47896841e-01 -1.76491678e-01 -4.14292365e-02
-1.67847908e+00 6.24147356e-01 2.55714148e-01 1.52236968e-01
-4.68188405e-01 -1.10241491e-02 4.28303212e-01 8.79548371e-01
6.81831002e-01 7.08546281e-01 -7.02228367e-01 -5.06169796e-01
-1.05052233e+00 -6.85609356e-02 8.06658387e-01 1.11972380e+00
5.97075880e-01 4.90667522e-02 -4.67372149e-01 8.42392385e-01
4.82471287e-02 1.01294541e+00 6.82383597e-01 -8.31335247e-01
1.07119262e+00 1.91779718e-01 1.11505754e-01 -1.41999528e-01
9.19086114e-02 -5.63510239e-01 -6.67531013e-01 -3.86991143e-01
1.87289402e-01 -2.96829492e-01 -1.01574838e+00 1.87455249e+00
-1.61696635e-02 -3.26192111e-01 2.27799892e-01 6.61858857e-01
2.68408060e-01 8.73844504e-01 -2.99501389e-01 -3.26235563e-01
7.98557162e-01 -1.41445255e+00 -6.56722307e-01 -4.35450673e-01
6.91575348e-01 -1.03680992e+00 1.60055304e+00 8.21215194e-03
-1.60175467e+00 -3.67833942e-01 -8.35160673e-01 -3.62652004e-01
7.33566210e-02 4.88028795e-01 2.14970618e-01 2.10212886e-01
-1.18210030e+00 9.63225424e-01 -1.19156420e+00 -1.87543780e-01
2.62042612e-01 4.50682849e-01 -3.77202034e-01 -2.82179296e-01
-1.20806491e+00 1.32989776e+00 8.99829492e-02 2.41023377e-02
-9.92046118e-01 -5.12669921e-01 -6.59400463e-01 -9.79435071e-02
6.27961010e-02 -1.06529450e+00 1.59479415e+00 -1.21494246e+00
-2.23839688e+00 7.45514631e-01 -5.75472116e-01 -5.07308662e-01
9.79648709e-01 -3.67309034e-01 -1.00056820e-01 -2.26774484e-01
1.28629595e-01 6.52034581e-01 6.23113215e-01 -7.35822022e-01
-1.93831488e-01 -1.16666339e-01 -2.40691811e-01 4.98316884e-01
-2.16986805e-01 2.02603206e-01 -4.94584680e-01 -6.45403087e-01
-7.85476118e-02 -1.07840109e+00 -1.43491760e-01 -3.37841570e-01
-5.26012421e-01 -5.39185852e-02 1.95452839e-01 -1.14709342e+00
8.55174661e-01 -1.56437218e+00 7.71098018e-01 -1.86397344e-01
-2.65999377e-01 2.45716289e-01 -6.40025616e-01 6.98124945e-01
3.89117837e-01 8.54981616e-02 -3.85444015e-01 -8.83398294e-01
2.88456082e-02 1.98332235e-01 -6.31440341e-01 1.76955849e-01
3.97591680e-01 1.39432466e+00 -8.45378339e-01 -1.98027760e-01
-2.09679693e-01 4.15563524e-01 -5.00938714e-01 3.80610466e-01
-5.08550406e-01 5.78305781e-01 -3.78996916e-02 4.61348146e-01
2.62953967e-01 -2.49951318e-01 3.17657053e-01 3.88086647e-01
2.84951068e-02 1.19386613e+00 -8.44008774e-02 1.93117464e+00
-9.29907024e-01 6.94471180e-01 -2.41936296e-01 -6.31470442e-01
8.40505481e-01 4.29710448e-01 -3.90853621e-02 -8.55445921e-01
-4.59780954e-02 6.16061509e-01 -8.84481594e-02 -1.15683982e-02
4.22657043e-01 -2.85257459e-01 5.92022650e-02 7.73524880e-01
1.06820256e-01 -5.67597859e-02 -8.57661366e-02 -1.47682413e-01
9.60562527e-01 4.77671891e-01 -7.50394985e-02 -1.45530224e-01
1.31813914e-01 1.03152968e-01 4.12028670e-01 5.14089584e-01
2.66821474e-01 6.02032602e-01 3.46984267e-01 -1.69306546e-01
-1.58697236e+00 -1.17400467e+00 2.60233521e-01 1.05739677e+00
-4.57960874e-01 -2.67711461e-01 -7.84302235e-01 -8.69027555e-01
-1.46263048e-01 1.05120194e+00 -3.12045217e-01 -1.71789750e-01
-1.08779168e+00 -4.50046867e-01 7.74340987e-01 4.33368355e-01
2.05315918e-01 -8.92171860e-01 -2.26262771e-02 2.95981973e-01
-6.14450634e-01 -1.01227701e+00 -9.03704524e-01 2.10595191e-01
-1.39326620e+00 -2.20401421e-01 -9.33724165e-01 -7.59692132e-01
6.48690045e-01 4.44884691e-03 1.46460140e+00 -2.89060503e-01
4.06769067e-01 -3.47312242e-01 5.97019587e-03 -1.31778652e-02
-1.06056559e+00 7.26321876e-01 1.98835239e-01 -3.66481632e-01
2.06413448e-01 -4.25969988e-01 -2.38090917e-01 3.17559749e-01
-4.68418777e-01 6.10194445e-01 9.35437799e-01 1.19747150e+00
6.56107187e-01 -9.75376785e-01 4.06472653e-01 -7.11314559e-01
7.78007865e-01 -2.69497335e-01 -4.71479297e-01 4.15842503e-01
-8.14313412e-01 5.49154341e-01 9.43714261e-01 -6.21172071e-01
-7.16706157e-01 -3.06509525e-01 -1.60538465e-01 -4.65553731e-01
3.00243765e-01 4.04402763e-01 -8.54416043e-02 3.73910218e-01
6.44083679e-01 6.01384640e-01 1.03051983e-01 -6.72495544e-01
5.68235338e-01 8.99318755e-01 4.20523435e-01 -6.76631808e-01
6.60995483e-01 -3.91581580e-02 -1.92309171e-01 -1.86739594e-01
-6.23499274e-01 1.18230969e-01 -6.63683295e-01 3.87791127e-01
5.37961602e-01 -9.26682234e-01 -9.61630568e-02 4.34679687e-02
-1.35257924e+00 -7.10937798e-01 -2.66321838e-01 7.36402154e-01
-9.09220576e-01 5.16287610e-02 -9.89037633e-01 -3.29975367e-01
-8.03451002e-01 -1.27093530e+00 1.11415577e+00 -5.22626281e-01
-2.84225196e-01 -8.99583757e-01 1.50488034e-01 3.94060403e-01
6.59856260e-01 -2.96611995e-01 1.14298511e+00 -6.25684679e-01
-6.36322856e-01 -2.09457085e-01 -7.32881129e-02 5.24965525e-01
2.10786626e-01 -2.98338801e-01 -5.12924314e-01 -6.22494459e-01
-4.01192717e-02 -5.02914369e-01 7.11223900e-01 9.48216990e-02
5.57855546e-01 -7.02683985e-01 -1.10101208e-01 6.57567084e-01
1.10126281e+00 -2.60760337e-01 3.94251049e-01 2.95041323e-01
6.08868361e-01 3.95874798e-01 2.61368454e-01 -1.24060079e-01
3.79116654e-01 8.68785739e-01 -6.19668253e-02 -1.28785998e-01
-3.73476565e-01 -8.21159899e-01 9.37692225e-01 1.35503006e+00
-2.72542681e-03 -3.10493737e-01 -9.50562716e-01 4.52245593e-01
-1.67893255e+00 -6.53594673e-01 3.56412143e-01 2.40547371e+00
1.27485991e+00 2.96335995e-01 4.27963361e-02 -5.79077184e-01
6.32151484e-01 -3.44456472e-02 -7.20695555e-01 -6.20765209e-01
-1.86651632e-01 2.93241858e-01 7.57618189e-01 8.09991241e-01
-5.59399962e-01 1.27218485e+00 7.63860989e+00 6.09546363e-01
-1.31371832e+00 4.83322144e-01 6.18132889e-01 -4.05677140e-01
-7.39317775e-01 9.38470438e-02 -6.76675260e-01 4.51888770e-01
1.63916814e+00 -1.84463784e-01 9.77938294e-01 3.98555994e-01
4.04528826e-01 5.87604821e-01 -1.49225008e+00 5.53689778e-01
-1.91703126e-01 -1.44645548e+00 3.39360237e-01 8.25885013e-02
9.19715106e-01 6.85571790e-01 2.53868222e-01 4.72914100e-01
5.78792930e-01 -9.88115668e-01 7.30997920e-01 3.46590281e-01
1.26875794e+00 -5.53059220e-01 5.27933776e-01 7.34681427e-01
-5.16706526e-01 3.31508309e-01 -4.07657504e-01 -9.37735960e-02
1.92634016e-01 1.75187558e-01 -1.25140393e+00 5.64677954e-01
-3.13417315e-02 5.66303551e-01 -1.95269659e-01 3.11305225e-01
-4.51488554e-01 9.36867297e-01 -2.72858530e-01 -9.85702351e-02
3.86800230e-01 -2.51437992e-01 5.44469059e-01 1.19815373e+00
4.70612317e-01 -5.18343389e-01 -3.69241908e-02 1.01542032e+00
-7.47237921e-01 9.30609554e-02 -5.53554654e-01 -2.59414792e-01
4.17601049e-01 6.23535037e-01 -2.78810523e-02 -5.09421647e-01
-3.51576775e-01 1.64531505e+00 6.07638955e-01 4.06154960e-01
-6.11032605e-01 -5.72709516e-02 5.93961954e-01 1.54624015e-01
6.53694272e-02 -4.25295144e-01 -3.58115762e-01 -1.47907937e+00
3.49095434e-01 -1.25348151e+00 -1.71796858e-01 -5.19503951e-01
-1.10833037e+00 8.36000919e-01 -2.43131205e-01 -1.18349016e+00
-8.57165039e-01 -3.60034078e-01 -3.31713825e-01 1.43107784e+00
-1.43118310e+00 -1.41342664e+00 6.89454734e-01 8.39818716e-02
9.63176370e-01 -3.38043749e-01 9.62785184e-01 2.04413772e-01
-5.48032582e-01 1.21121132e+00 6.29602790e-01 -8.43053907e-02
8.81813288e-01 -1.12124467e+00 1.27248478e+00 9.65845525e-01
2.61757761e-01 8.69844913e-01 5.76257586e-01 -5.66974521e-01
-1.78024316e+00 -1.20488524e+00 1.65918827e+00 -7.50330925e-01
5.01348794e-01 -5.51850736e-01 -6.69457972e-01 9.51979876e-01
3.98524910e-01 -2.35951692e-01 3.82409960e-01 1.72137335e-01
-4.76869643e-01 3.54825437e-01 -9.50392604e-01 8.52651834e-01
1.11126876e+00 -9.38552976e-01 -7.04452395e-01 6.48034096e-01
1.16480505e+00 -5.72748780e-01 -8.78716707e-01 3.61200035e-01
4.27292317e-01 -1.99011922e-01 7.00381637e-01 -1.09163690e+00
8.60250473e-01 1.93752989e-01 -3.43165606e-01 -1.63119268e+00
-1.85093030e-01 -1.13183331e+00 -1.92307115e-01 8.23978126e-01
1.10772228e+00 -7.52625167e-01 7.27632701e-01 4.31777894e-01
-2.51063854e-01 -1.00474405e+00 -9.78653073e-01 -1.12607372e+00
8.02563012e-01 -2.73189768e-02 6.84155107e-01 7.92425096e-01
2.96099205e-02 6.91952765e-01 -6.51519358e-01 -2.11785346e-01
2.89185286e-01 1.57042623e-01 8.52271140e-01 -4.68528420e-01
-8.40537548e-01 -5.12810051e-01 2.53078848e-01 -1.60133660e+00
3.94726545e-01 -1.42767262e+00 1.11875078e-03 -1.50537276e+00
2.72054315e-01 -1.36059672e-01 -2.12844551e-01 5.03242314e-01
-2.87228435e-01 3.16140801e-01 2.78134525e-01 7.44731784e-01
-2.96159834e-01 7.64167130e-01 1.33270597e+00 -1.76038548e-01
-8.81423429e-02 9.21712965e-02 -4.28301215e-01 2.75530100e-01
6.14897311e-01 -6.16587281e-01 -1.19895376e-01 -1.27721953e+00
7.30145797e-02 4.29588109e-01 -1.87944189e-01 -4.77982789e-01
-6.68669567e-02 -2.69510746e-01 2.05740780e-01 -3.65150243e-01
2.40012318e-01 -3.44970912e-01 1.30221069e-01 5.24927855e-01
-9.59331393e-01 6.44918203e-01 1.08713940e-01 2.00516880e-01
-1.61019087e-01 6.96548373e-02 6.89769387e-01 -7.73684159e-02
1.64652780e-01 9.86500159e-02 -1.43703148e-01 8.83442387e-02
3.69134128e-01 1.47249952e-01 -9.06326249e-02 -3.41779858e-01
-4.02183592e-01 -7.46131828e-03 7.94130683e-01 6.20992243e-01
3.96501631e-01 -1.61150849e+00 -1.19924903e+00 2.26144865e-01
-1.39589794e-02 -3.42191458e-01 -4.57819819e-01 8.38082969e-01
-4.24164891e-01 6.71508431e-01 -1.39661476e-01 -4.11809891e-01
-1.05031335e+00 3.41988206e-01 3.50882798e-01 -6.23036385e-01
-5.94090462e-01 6.72505438e-01 -4.20971274e-01 -1.08406806e+00
-5.70052601e-02 -2.06969202e-01 6.76755965e-01 -6.19775116e-01
3.55584651e-01 2.61517972e-01 3.99133325e-01 -5.51103532e-01
-1.37548223e-01 1.13857657e-01 -2.52062023e-01 -7.99221098e-01
1.11165130e+00 -1.56521961e-01 -6.16345927e-03 6.38497233e-01
1.46271420e+00 1.27378916e-02 -1.20349491e+00 -7.56859601e-01
1.01695433e-01 -4.39149201e-01 -1.11371122e-01 -1.22141194e+00
-4.55545783e-01 1.06162965e+00 3.22161227e-01 -5.69424689e-01
6.85409069e-01 -2.13086113e-01 1.28304625e+00 7.24636137e-01
4.35018480e-01 -8.91294062e-01 -1.25407040e-01 8.61880362e-01
9.29985821e-01 -1.15587306e+00 -2.45841071e-01 1.11938231e-01
-4.56365615e-01 9.30087566e-01 1.86035022e-01 2.99934633e-02
-3.11659109e-02 1.31131321e-01 2.73167074e-01 4.12130892e-01
-1.22214866e+00 1.86388597e-01 3.97420466e-01 2.71495938e-01
7.62598693e-01 3.19354028e-01 -3.98870707e-01 1.15627594e-01
-4.34247136e-01 1.06953532e-02 2.25176051e-01 6.62681818e-01
-1.15080334e-01 -1.55507636e+00 -9.42981020e-02 4.02587473e-01
-7.49072015e-01 -5.38157165e-01 -6.62552178e-01 3.79815131e-01
-5.47055244e-01 7.87245810e-01 -4.62961122e-02 -4.78342712e-01
3.62886757e-01 3.13854039e-01 6.64168417e-01 -5.56492329e-01
-6.96435213e-01 5.78937493e-02 2.31086358e-01 -4.27913427e-01
2.20960721e-01 -7.20948517e-01 -9.01294887e-01 -4.63239819e-01
-3.07136476e-01 1.06175028e-01 9.88473594e-01 8.60304534e-01
7.21227884e-01 2.53983200e-01 8.90157580e-01 -8.78258467e-01
-1.45238590e+00 -1.28887153e+00 4.43448275e-01 3.69149327e-01
3.89668733e-01 9.30709590e-04 -1.96820036e-01 -7.56469965e-02] | [11.660714149475098, 10.222047805786133] |
b81feb16-f0b3-4376-8d6f-47f42a8e2099 | associative-learning-mechanism-for-drug | 2205.15364 | null | https://arxiv.org/abs/2205.15364v4 | https://arxiv.org/pdf/2205.15364v4.pdf | Associative Learning Mechanism for Drug-Target Interaction Prediction | As a necessary process in drug development, finding a drug compound that can selectively bind to a specific protein is highly challenging and costly. Drug-target affinity (DTA), which represents the strength of drug-target interaction (DTI), has played an important role in the DTI prediction task over the past decade. Although deep learning has been applied to DTA-related research, existing solutions ignore fundamental correlations between molecular substructures in molecular representation learning of drug compound molecules/protein targets. Moreover, traditional methods lack the interpretability of the DTA prediction process. This results in missing feature information of intermolecular interactions, thereby affecting prediction performance. Therefore, this paper proposes a DTA prediction method with interactive learning and an autoencoder mechanism. The proposed model enhances the corresponding ability to capture the feature information of a single molecular sequence by the drug/protein molecular representation learning module and supplements the information interaction between molecular sequence pairs by the interactive information learning module. The DTA value prediction module fuses the drug-target pair interaction information to output the predicted value of DTA. Additionally, this paper theoretically proves that the proposed method maximizes evidence lower bound (ELBO) for the joint distribution of the DTA prediction model, which enhances the consistency of the probability distribution between the actual value and the predicted value. The experimental results confirm mutual transformer-drug target affinity (MT-DTA) achieves better performance than other comparative methods. | ['Baisen Cong', 'Neal Mazur', 'Guanqiu Qi', 'Zheng Yao', 'Zhiqin Zhu'] | 2022-05-24 | null | null | null | null | ['value-prediction'] | ['computer-code'] | [ 6.60816878e-02 -2.67713964e-01 -6.31758988e-01 -3.99402469e-01
-1.90681905e-01 -1.99690923e-01 2.31300414e-01 3.23995769e-01
-9.69222113e-02 1.05130661e+00 3.40685174e-02 -4.83369499e-01
-6.19541407e-01 -9.48101044e-01 -7.47187555e-01 -1.17180002e+00
-9.91246700e-02 4.55303937e-01 -4.64440770e-02 -1.26599997e-01
1.52625188e-01 6.51785135e-01 -1.13645887e+00 5.98245263e-01
1.28701127e+00 8.32037091e-01 5.53950429e-01 7.94915631e-02
-3.39675963e-01 7.84325004e-01 -5.60390353e-01 -2.21702874e-01
-1.42162800e-01 -5.11126697e-01 -6.20268822e-01 -4.35551077e-01
-1.82229817e-01 -3.84045720e-01 -2.58023858e-01 9.51722443e-01
7.16270983e-01 -7.71104842e-02 1.04300737e+00 -9.40383375e-01
-6.63661003e-01 4.56854075e-01 -3.68653983e-01 1.27942199e-02
2.95673937e-01 8.93168151e-02 9.76437449e-01 -8.77183616e-01
5.46094418e-01 1.21647871e+00 4.72939074e-01 2.38543004e-01
-9.28449869e-01 -6.71405196e-01 -4.30376120e-02 5.60230136e-01
-1.28178978e+00 4.37229782e-01 6.20868385e-01 -6.13318503e-01
1.19339728e+00 3.79857011e-02 7.06097543e-01 8.06355655e-01
7.47906983e-01 8.40852380e-01 8.35813761e-01 -1.57802373e-01
1.89721640e-02 1.35231763e-01 1.74610198e-01 4.52231526e-01
1.97441399e-01 4.28125679e-01 -1.99935645e-01 -2.17052519e-01
6.85153782e-01 4.03114140e-01 -2.68352449e-01 -8.60528126e-02
-8.68567050e-01 9.11516488e-01 5.69654882e-01 4.74447191e-01
-5.04935801e-01 -4.91762757e-01 1.87353402e-01 1.36453569e-01
-1.10991083e-01 2.53288865e-01 -7.26505816e-01 9.68532264e-02
-3.18086982e-01 1.46438897e-01 7.03103185e-01 5.44045269e-01
6.61303401e-01 -6.97013363e-02 -8.62935036e-02 6.82312965e-01
5.41616857e-01 4.54794526e-01 6.50759041e-01 -1.24311976e-01
7.87331164e-03 8.41031194e-01 3.91713567e-02 -1.03192246e+00
-4.75345820e-01 -4.76457387e-01 -8.37981343e-01 1.69407114e-01
2.75381416e-01 -7.34057724e-02 -7.09974229e-01 1.66876709e+00
5.60538828e-01 2.07048431e-02 3.91780972e-01 8.68821800e-01
9.49591100e-01 1.05660760e+00 5.85449934e-01 -6.08295858e-01
1.02205670e+00 -6.11150980e-01 -9.12150502e-01 3.83993983e-01
8.29401314e-01 -8.70200872e-01 3.95209461e-01 2.70501584e-01
-4.70844269e-01 -6.16651237e-01 -1.14648294e+00 1.42451897e-01
-4.81387556e-01 1.38848528e-01 8.81208181e-01 1.83701724e-01
-3.52683440e-02 8.45917225e-01 -5.16540349e-01 -1.55212230e-03
3.02247852e-01 9.42264140e-01 -4.21094000e-01 4.79140617e-02
-1.67098379e+00 1.22337008e+00 7.53327370e-01 -5.72330803e-02
-6.94869876e-01 -7.75698185e-01 -4.76971209e-01 1.32564038e-01
-8.84952843e-02 -6.47656679e-01 7.57590771e-01 -9.47870791e-01
-1.51387107e+00 7.97288790e-02 -4.37383205e-02 -2.84425944e-01
-2.90414300e-02 -1.67320445e-01 -5.38660765e-01 -1.14983954e-01
-2.43843764e-01 4.40261841e-01 2.96199262e-01 -8.71168494e-01
-3.61754179e-01 -3.64428729e-01 -2.13897184e-01 3.23332816e-01
-1.42730713e-01 -3.42601240e-01 1.37402788e-01 -4.30241019e-01
-1.97935820e-01 -4.83347356e-01 -1.01136491e-01 -3.74852061e-01
-1.45172223e-01 -6.65196836e-01 7.92867422e-01 -5.45708418e-01
1.37883437e+00 -2.01715279e+00 3.36044252e-01 4.92496490e-01
1.10265389e-01 6.00954890e-01 -1.08258113e-01 8.07301104e-01
-5.79503834e-01 -7.12516531e-02 -1.40274316e-01 9.05324101e-01
-3.99957299e-01 2.31779978e-01 -1.56396568e-01 2.29852274e-01
2.27851048e-01 7.06673682e-01 -8.12370181e-01 -5.12431026e-01
4.23470080e-01 6.89216077e-01 -4.16285932e-01 5.26912510e-01
-5.12079358e-01 3.61522198e-01 -8.84245038e-01 5.27142346e-01
9.58052516e-01 -3.90005231e-01 6.32409215e-01 -5.36816895e-01
-2.13978849e-02 2.69848034e-02 -8.26049864e-01 1.15482330e+00
-3.96344736e-02 2.91845296e-02 -7.33380854e-01 -1.11245012e+00
1.12495840e+00 4.45921272e-01 8.95326614e-01 -6.61015272e-01
3.02315086e-01 2.64938891e-01 5.88882208e-01 -7.89093733e-01
-3.66676927e-01 -2.93078691e-01 6.34929419e-01 2.28506653e-03
1.43399134e-01 4.69445348e-01 -2.30730593e-01 -1.35496914e-01
7.17603147e-01 2.26045772e-01 6.71709001e-01 -1.13721624e-01
7.66756177e-01 -1.31951705e-01 7.04193830e-01 2.62414813e-01
7.48767480e-02 -2.48321053e-02 3.92242700e-01 -6.80716336e-01
-1.06591415e+00 -8.06848526e-01 -7.50352263e-01 6.15184546e-01
4.33997095e-01 -1.93870172e-01 -6.40278697e-01 -6.30743802e-01
2.78504461e-01 4.14456069e-01 -5.91743827e-01 -5.12312889e-01
-2.37308234e-01 -1.19928217e+00 5.37008494e-02 3.52818549e-01
3.31923962e-01 -9.97681499e-01 2.83263564e-01 5.34732103e-01
1.74994156e-01 -4.26233023e-01 -1.19095244e-01 2.35363171e-01
-8.60188127e-01 -1.30140519e+00 -6.00471497e-01 -8.13660383e-01
4.37364370e-01 -4.32865359e-02 6.35446310e-01 7.82957971e-02
-1.37634933e-01 -4.30612296e-01 -2.12306798e-01 -5.18979073e-01
-4.77610707e-01 -2.91144341e-01 1.37065008e-01 -3.46550159e-02
9.03642416e-01 -6.15551829e-01 -8.10819745e-01 4.40427035e-01
-7.46692121e-01 3.20743956e-02 9.77266192e-01 1.14872658e+00
7.03876257e-01 1.55777156e-01 8.64713609e-01 -6.07334137e-01
6.20205760e-01 -6.76068604e-01 -5.60111880e-01 3.58774543e-01
-8.27609479e-01 2.23124802e-01 6.62794232e-01 -6.15372121e-01
-1.00329232e+00 4.26424891e-02 -5.51994026e-01 -1.79238454e-01
-3.53632830e-02 9.67422724e-01 -6.45673037e-01 -8.53784531e-02
4.21588540e-01 4.68763471e-01 2.01158598e-01 -3.46770585e-01
-1.99716464e-01 7.40752518e-01 -1.20992370e-01 -4.36620921e-01
8.72371718e-02 -2.10460588e-01 2.34570876e-01 -5.61557353e-01
-4.48978543e-01 -3.68721604e-01 -4.06804562e-01 -1.37843341e-01
7.95805216e-01 -8.74083400e-01 -1.43170714e+00 3.03770900e-01
-1.25943696e+00 2.05900088e-01 3.61077726e-01 1.24349391e+00
-3.73076499e-01 6.90770268e-01 -6.06675863e-01 -5.70331573e-01
-4.24701601e-01 -1.41172481e+00 5.96614838e-01 3.82618338e-01
-9.94724184e-02 -1.04145849e+00 1.61888644e-01 2.59072900e-01
1.32128336e-02 2.43232414e-01 1.43207002e+00 -1.02004385e+00
-5.97113311e-01 -3.49867702e-01 -1.47554323e-01 3.27213705e-01
2.60277152e-01 4.29034978e-02 -7.21866310e-01 -3.14935707e-02
-1.58622246e-02 -1.32736135e-02 6.06289208e-01 7.81026721e-01
1.15534425e+00 -3.38390857e-01 -5.48131287e-01 1.88091069e-01
1.40604389e+00 1.07367277e+00 7.90608525e-01 1.68711141e-01
6.84418976e-01 2.92402834e-01 8.63151968e-01 4.92294312e-01
-2.86921561e-02 6.74679518e-01 5.35534799e-01 -1.25826970e-01
1.62554502e-01 -2.82669038e-01 3.10509950e-01 7.47014046e-01
-2.14772254e-01 -4.13264036e-01 -7.01125860e-01 1.66669980e-01
-2.04545641e+00 -1.18564832e+00 -1.08928770e-01 2.11024356e+00
1.31982672e+00 -1.62483066e-01 -3.35994177e-02 -5.91690280e-02
7.17946768e-01 -4.38763946e-01 -9.31034386e-01 -3.67828101e-01
-1.55559167e-01 7.33515397e-02 2.29558259e-01 7.04761207e-01
-8.46508741e-01 6.53983533e-01 6.19030476e+00 1.30270493e+00
-1.18905258e+00 -3.61391127e-01 6.01967454e-01 5.20379424e-01
-3.55170310e-01 -2.65539736e-01 -7.61760175e-01 8.23833942e-01
9.07231092e-01 -1.75717697e-01 6.82763159e-02 1.08004892e+00
2.94408709e-01 9.15934369e-02 -1.14249563e+00 8.40768814e-01
-4.38629419e-01 -1.43349731e+00 6.14137888e-01 3.41294438e-01
5.56461990e-01 -3.63145202e-01 1.74081147e-01 6.44135848e-02
1.50795400e-01 -1.21971452e+00 -2.36166134e-01 7.59784579e-01
3.93485487e-01 -9.98129308e-01 1.22170651e+00 3.59773815e-01
-8.22557092e-01 -7.52991363e-02 -5.84817886e-01 -1.90113112e-02
-2.86789000e-01 8.22840393e-01 -1.15783024e+00 7.94241726e-01
1.88115612e-01 9.96137083e-01 -2.16614399e-02 1.08410323e+00
-5.01402700e-03 4.02651995e-01 -2.01583371e-01 -3.99351060e-01
1.79375485e-01 -6.04459584e-01 2.24579513e-01 9.48835492e-01
2.28981972e-01 3.08562011e-01 1.71393812e-01 6.45203352e-01
-4.40063067e-02 6.07559085e-01 -5.51574707e-01 -4.07100946e-01
4.63920057e-01 7.79796302e-01 -1.00149900e-01 -3.83633018e-01
-4.15933102e-01 5.97809434e-01 5.77259585e-02 2.87365258e-01
-9.11652684e-01 -5.06274760e-01 6.32489204e-01 -5.26938438e-02
1.36878371e-01 2.82665044e-01 -2.17918400e-03 -6.86641812e-01
-3.13692987e-01 -9.94598806e-01 3.83260846e-01 -5.35564840e-01
-1.50530756e+00 3.23598534e-01 -2.38898113e-01 -1.35389519e+00
1.13310553e-02 -9.66539264e-01 -3.99256110e-01 1.15977204e+00
-1.20815611e+00 -1.02827978e+00 2.42967770e-01 4.44273740e-01
3.23157847e-01 -3.95636290e-01 1.00712681e+00 5.12445152e-01
-6.28529429e-01 5.06560624e-01 6.77940369e-01 -2.88635958e-02
7.28456259e-01 -9.93235588e-01 -4.29750323e-01 3.87079120e-02
-4.21207756e-01 9.23230469e-01 6.42968595e-01 -8.85754287e-01
-1.34349763e+00 -7.86464989e-01 7.04525650e-01 -1.68395668e-01
4.34485704e-01 3.73335510e-01 -1.01716375e+00 2.67063230e-01
1.38209939e-01 -3.76963735e-01 1.24929857e+00 1.11807495e-01
-1.27278358e-01 -1.39504910e-01 -1.04553688e+00 3.44907433e-01
4.19701159e-01 -4.57210898e-01 -5.19165397e-01 6.24532819e-01
6.72966957e-01 -2.52911925e-01 -1.48915434e+00 8.07371616e-01
8.30072165e-01 -7.38318622e-01 1.15559053e+00 -1.00957346e+00
4.41587538e-01 -6.69882536e-01 -3.48468646e-02 -9.00785029e-01
-5.49231291e-01 1.68062776e-01 -1.63439125e-01 7.08803356e-01
5.40584207e-01 -4.12717551e-01 6.30633652e-01 5.17901182e-01
-3.40015553e-02 -1.35120368e+00 -8.73059928e-01 -6.63954973e-01
2.41038010e-01 1.42534375e-01 8.32253039e-01 1.10957634e+00
3.26446861e-01 5.69109142e-01 -4.91947114e-01 2.74137557e-01
3.22254539e-01 2.70741671e-01 3.31923604e-01 -1.23493171e+00
-4.19143885e-01 -4.30242151e-01 -5.65374732e-01 -9.67856407e-01
-7.44024366e-02 -9.76031184e-01 -5.71152091e-01 -1.48916280e+00
4.74674851e-01 -1.31558523e-01 -7.02281177e-01 3.07468176e-01
-2.40188599e-01 -5.67806900e-01 -3.35693359e-01 3.19497943e-01
-2.12958813e-01 7.40020573e-01 1.59419000e+00 -4.72374290e-01
-4.01962578e-01 -5.19259050e-02 -3.91961426e-01 4.61061001e-01
6.97832644e-01 -2.73150980e-01 -5.33867955e-01 1.13463640e-01
1.46022514e-01 3.03270310e-01 -1.47370785e-01 -7.47171760e-01
5.98421246e-02 -5.15137553e-01 9.49477494e-01 -6.27040923e-01
1.41197175e-01 -1.11338913e+00 3.78284782e-01 7.86471128e-01
-2.85806954e-01 -4.89176393e-01 1.83523014e-01 6.53748512e-01
-3.37711006e-01 -2.71694243e-01 6.94972515e-01 1.62370242e-02
-5.10115266e-01 6.17712498e-01 -4.01746601e-01 -6.88993156e-01
1.05197573e+00 -3.58677179e-01 -7.93816335e-03 4.64641210e-03
-8.79919231e-01 2.21154988e-02 -4.39502746e-02 2.22909868e-01
7.52625883e-01 -1.36044478e+00 -3.52657914e-01 9.33013707e-02
6.18303865e-02 -3.13390613e-01 5.17806768e-01 5.96093893e-01
-4.58604395e-01 5.72837949e-01 -3.15822721e-01 -4.88785595e-01
-1.31680918e+00 8.51810455e-01 6.03110552e-01 -2.59800971e-01
-2.40690544e-01 5.30624807e-01 6.26072586e-01 -2.09752366e-01
2.78841518e-02 1.38970405e-01 -6.31837904e-01 -1.41999871e-01
5.08550704e-01 2.67166011e-02 -1.48094594e-01 -5.16166210e-01
-3.95760268e-01 6.54451370e-01 -6.47974908e-01 6.39740884e-01
1.29405046e+00 3.25577855e-01 -1.26273990e-01 -3.06334104e-02
1.38236511e+00 -3.08056474e-01 -9.04517472e-01 -8.68401676e-02
-1.63554013e-01 -2.49676496e-01 1.45188272e-01 -1.40927780e+00
-5.66002846e-01 8.92544627e-01 9.23457980e-01 -2.69168407e-01
8.22599649e-01 -1.51428774e-01 8.21217000e-01 5.90159297e-01
1.29614715e-02 -9.67901886e-01 2.39644572e-02 2.46072978e-01
8.31957579e-01 -1.31695795e+00 2.09742993e-01 -4.00741220e-01
-5.99577844e-01 1.31440234e+00 6.40904546e-01 3.45789015e-01
7.93867111e-01 -4.73793037e-02 -1.09627306e-01 -2.62683362e-01
-5.04001856e-01 1.39986783e-01 2.63367325e-01 5.20499051e-01
7.58130550e-01 1.26279280e-01 -7.72196591e-01 8.16907048e-01
2.86207080e-01 1.96155146e-01 -1.52335554e-01 6.03979826e-01
-6.02816463e-01 -1.58163881e+00 -1.12315975e-01 2.81292677e-01
-4.27246332e-01 -2.43923396e-01 -4.26946044e-01 7.19005048e-01
4.05473113e-01 7.17376888e-01 -2.90062517e-01 -6.48349345e-01
3.55892666e-02 1.33951247e-01 4.12961215e-01 -4.86488119e-02
-2.35162735e-01 2.38696471e-01 -2.71768719e-01 -1.72807589e-01
-4.21671152e-01 -1.86248317e-01 -1.68683589e+00 -4.43129539e-01
-6.65496826e-01 5.93142211e-01 5.35163641e-01 9.72138166e-01
7.53992200e-01 5.55441022e-01 8.02533448e-01 -1.46436855e-01
-2.03922346e-01 -8.99605870e-01 -5.65723181e-01 3.15374464e-01
9.40997899e-02 -8.31339419e-01 6.98424056e-02 -8.66229236e-02] | [5.139306545257568, 5.794998645782471] |
11acad6b-94d1-42c0-a990-ad0ea673f63d | hdr-video-reconstruction-a-coarse-to-fine | 2103.14943 | null | https://arxiv.org/abs/2103.14943v2 | https://arxiv.org/pdf/2103.14943v2.pdf | HDR Video Reconstruction: A Coarse-to-fine Network and A Real-world Benchmark Dataset | High dynamic range (HDR) video reconstruction from sequences captured with alternating exposures is a very challenging problem. Existing methods often align low dynamic range (LDR) input sequence in the image space using optical flow, and then merge the aligned images to produce HDR output. However, accurate alignment and fusion in the image space are difficult due to the missing details in the over-exposed regions and noise in the under-exposed regions, resulting in unpleasing ghosting artifacts. To enable more accurate alignment and HDR fusion, we introduce a coarse-to-fine deep learning framework for HDR video reconstruction. Firstly, we perform coarse alignment and pixel blending in the image space to estimate the coarse HDR video. Secondly, we conduct more sophisticated alignment and temporal fusion in the feature space of the coarse HDR video to produce better reconstruction. Considering the fact that there is no publicly available dataset for quantitative and comprehensive evaluation of HDR video reconstruction methods, we collect such a benchmark dataset, which contains $97$ sequences of static scenes and 184 testing pairs of dynamic scenes. Extensive experiments show that our method outperforms previous state-of-the-art methods. Our dataset, code and model will be made publicly available. | ['Lei Zhang', 'Kwan-Yee K. Wong', 'Zhetong Liang', 'Shi Guo', 'Chaofeng Chen', 'GuanYing Chen'] | 2021-03-27 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Chen_HDR_Video_Reconstruction_A_Coarse-To-Fine_Network_and_a_Real-World_Benchmark_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Chen_HDR_Video_Reconstruction_A_Coarse-To-Fine_Network_and_a_Real-World_Benchmark_ICCV_2021_paper.pdf | iccv-2021-1 | ['video-reconstruction', 'hdr-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 1.80638611e-01 -7.99374223e-01 6.45006597e-02 -1.62663817e-01
-6.38511181e-01 -2.39117891e-01 2.67415941e-01 -5.88976920e-01
-3.10126185e-01 6.62018955e-01 4.64244634e-01 8.95315334e-02
9.85631943e-02 -6.75228238e-01 -7.35582292e-01 -7.52371550e-01
9.15067457e-03 -9.11671892e-02 2.82786727e-01 -2.67239571e-01
9.69121158e-02 3.98631036e-01 -1.50115883e+00 3.85905147e-01
7.68168211e-01 9.79708433e-01 5.40058017e-01 7.39197135e-01
2.06823960e-01 1.12944233e+00 -2.86747873e-01 -3.22852254e-01
7.43598342e-01 -5.67212880e-01 -6.92039669e-01 2.72705257e-01
6.61787570e-01 -1.03119814e+00 -1.03196764e+00 1.28946602e+00
6.26541615e-01 2.80107617e-01 1.47329867e-01 -9.74797547e-01
-6.73381686e-01 2.72220820e-01 -1.00298166e+00 4.61593062e-01
8.16039562e-01 4.34584826e-01 7.18547642e-01 -8.86421025e-01
8.56694400e-01 1.33488166e+00 4.72918063e-01 2.14986876e-01
-1.09061515e+00 -9.49469984e-01 -9.58102643e-02 3.84950519e-01
-1.48563230e+00 -5.74142158e-01 8.85810137e-01 -2.65593171e-01
8.16303372e-01 1.66796207e-01 8.91705632e-01 9.46208656e-01
3.17243993e-01 5.40887117e-01 1.18990171e+00 -4.28010523e-02
-2.11190268e-01 -5.78692675e-01 -3.78346652e-01 7.20363557e-01
3.18605676e-02 5.69796681e-01 -6.50145352e-01 2.62960613e-01
1.35776949e+00 4.56414819e-01 -7.26101279e-01 -2.18729034e-01
-1.62435019e+00 4.47046161e-01 4.59769696e-01 2.05243915e-01
-3.71248722e-01 1.00177988e-01 2.55719781e-01 4.92507994e-01
2.71940291e-01 3.02750934e-02 1.46570420e-02 -2.09083796e-01
-1.05954075e+00 2.63983697e-01 4.77494359e-01 8.87099981e-01
8.85414541e-01 1.92908391e-01 -7.57643133e-02 8.95690680e-01
3.16601366e-01 5.32025874e-01 4.33931679e-01 -1.46919084e+00
6.84825063e-01 7.03330105e-03 1.09950706e-01 -1.30061984e+00
-5.96679859e-02 -3.22467424e-02 -1.25365961e+00 2.60374665e-01
2.66751945e-01 1.13644116e-01 -7.60491669e-01 1.32616174e+00
1.66928649e-01 6.20940268e-01 1.72210280e-02 1.49090505e+00
9.75391448e-01 9.81443286e-01 -1.49996281e-01 -5.68048418e-01
9.27648544e-01 -1.10481632e+00 -9.22764540e-01 -7.88267255e-02
7.53344744e-02 -1.02243304e+00 9.37514365e-01 3.54140043e-01
-1.33548045e+00 -9.58649278e-01 -1.20931756e+00 -4.58286077e-01
2.23780692e-01 -3.62142622e-01 3.33165437e-01 2.03235485e-02
-1.04671550e+00 6.60420239e-01 -5.33529222e-01 2.02513691e-02
2.34858662e-01 1.09327674e-01 -6.39043748e-01 -7.54725158e-01
-1.44435930e+00 6.92210555e-01 2.42330551e-01 2.50187308e-01
-1.04627550e+00 -8.45536411e-01 -1.02897012e+00 -2.52678424e-01
1.33486792e-01 -6.68803990e-01 8.09175789e-01 -9.54548419e-01
-1.37950408e+00 8.23765635e-01 7.98823833e-02 -2.83460438e-01
7.86416054e-01 -2.59841144e-01 -5.10634601e-01 2.38878608e-01
-4.78394777e-02 6.52735770e-01 1.05177915e+00 -1.27444577e+00
-6.92337871e-01 -1.25775799e-01 3.77984866e-02 4.91622746e-01
1.58047810e-01 -3.58791277e-02 -7.01191127e-01 -8.94916594e-01
2.30391324e-01 -7.99520910e-01 -2.30300039e-01 5.54531626e-02
-2.24144608e-01 5.28190970e-01 9.89723682e-01 -1.10805881e+00
1.33860505e+00 -2.27871227e+00 3.22520375e-01 -1.93972796e-01
5.83474278e-01 6.23923279e-02 -2.62955606e-01 4.71584722e-02
-2.65147626e-01 -3.94171864e-01 -1.26391754e-01 -1.64859280e-01
-3.26992214e-01 -6.44314103e-04 -5.42846322e-01 8.22120905e-01
-3.56217355e-01 7.16867566e-01 -1.01753914e+00 -7.27730453e-01
8.03807735e-01 5.92046082e-01 -5.27309537e-01 7.34238029e-01
2.44246930e-01 8.00558746e-01 -1.92484573e-01 7.57938445e-01
8.70908737e-01 -8.50412473e-02 -1.30430803e-01 -8.04959238e-01
-1.71728954e-01 -8.34165290e-02 -1.24976158e+00 1.94672978e+00
-4.18457031e-01 8.50418508e-01 -1.29346564e-01 -7.34949708e-01
9.07606125e-01 8.15047622e-02 1.03442538e+00 -1.06090677e+00
2.52657086e-02 2.59669036e-01 -3.11946236e-02 -7.09024847e-01
7.49109685e-01 -2.08108984e-02 3.99120450e-02 3.45541716e-01
-1.19547918e-01 -1.52981117e-01 1.87127650e-01 7.84447715e-02
9.69173729e-01 2.67358035e-01 4.08260971e-01 2.06265613e-01
5.67453980e-01 -4.35666829e-01 7.50210047e-01 4.15415615e-01
-4.71404046e-01 1.18618476e+00 3.94604132e-02 -8.23208690e-01
-1.47073281e+00 -1.23749554e+00 -2.28369720e-02 7.49792695e-01
7.25855827e-01 -2.13263914e-01 -4.24227238e-01 -1.70853168e-01
-2.61730403e-01 7.78648704e-02 -3.05571705e-01 -1.60377603e-02
-1.07905269e+00 -6.09181225e-01 1.37846500e-01 2.34777629e-01
1.15001774e+00 -1.11633968e+00 -4.80730742e-01 3.04486096e-01
-7.36396551e-01 -1.32406175e+00 -8.54232907e-01 -3.20449471e-01
-7.06537664e-01 -9.60058451e-01 -9.34329629e-01 -8.01178455e-01
3.29190671e-01 7.12768614e-01 1.25799382e+00 1.85694188e-01
-3.29506099e-01 -2.24066805e-02 -3.30493122e-01 4.78947788e-01
-3.90647113e-01 -4.70909864e-01 -2.99376510e-02 -8.82135406e-02
-2.02884581e-02 -6.21031046e-01 -9.38658416e-01 4.33799982e-01
-9.90137875e-01 5.15148580e-01 5.41487157e-01 9.48148012e-01
7.47282565e-01 1.64833218e-01 2.66962908e-02 -3.85081232e-01
1.59389421e-01 -2.79023767e-01 -5.49182475e-01 1.19565405e-01
-1.57356992e-01 -1.81317896e-01 6.10518277e-01 -3.72389942e-01
-1.04316306e+00 1.27791986e-01 -3.34674627e-01 -1.02613044e+00
6.57831952e-02 -1.54715836e-01 -2.79887527e-01 -1.57222778e-01
2.57628202e-01 5.04709482e-01 4.23954194e-03 -2.60811955e-01
3.19583476e-01 4.39050496e-01 9.04208302e-01 -3.50659996e-01
9.02760804e-01 5.53069651e-01 -1.66268364e-01 -5.61832309e-01
-7.18497038e-01 -3.81230861e-01 -5.19044459e-01 -5.45763135e-01
1.03869069e+00 -1.31590950e+00 -5.02358198e-01 6.61208391e-01
-9.76411283e-01 -3.50799680e-01 -1.27602637e-01 5.87278068e-01
-8.44782889e-01 6.15815580e-01 -9.56405044e-01 -1.28030837e-01
-4.33207661e-01 -1.61925459e+00 1.12275589e+00 2.51958907e-01
1.07591413e-01 -6.99753046e-01 2.72081316e-01 4.08758730e-01
4.19791102e-01 1.96810961e-01 3.91176671e-01 1.68192446e-01
-1.05574679e+00 2.12020263e-01 -4.49950308e-01 2.74171770e-01
1.59255669e-01 -8.02388042e-03 -7.06061006e-01 -5.60354292e-01
1.39431223e-01 -2.03365639e-01 8.01895380e-01 5.61429322e-01
1.28863931e+00 -2.77867019e-01 -8.36901814e-02 1.26037657e+00
1.59733045e+00 1.64050639e-01 1.18304217e+00 4.56808269e-01
1.14980257e+00 3.14177543e-01 9.75633204e-01 5.25071383e-01
3.95769626e-01 9.22824025e-01 2.73783684e-01 -3.34078908e-01
-5.42586803e-01 -2.89812833e-01 5.61004996e-01 9.18409705e-01
-3.08754772e-01 -1.21816292e-01 -4.15742934e-01 2.20432341e-01
-1.66372776e+00 -1.33661115e+00 1.99944809e-01 2.03281593e+00
9.48792040e-01 -2.07162902e-01 -1.09883910e-02 8.80369470e-02
8.33837688e-01 7.72018433e-01 -5.29714644e-01 6.74230158e-02
-2.86449641e-01 -1.79438859e-01 4.62799191e-01 5.67189872e-01
-1.02953064e+00 7.34346092e-01 6.24976254e+00 8.85678530e-01
-1.10542083e+00 2.89278533e-02 8.73751760e-01 -2.78515488e-01
-2.10849062e-01 -1.14350319e-01 -6.29507780e-01 8.21132362e-01
4.87786919e-01 -5.78091741e-02 8.41227710e-01 4.87668395e-01
2.13257834e-01 1.10568311e-02 -1.08302867e+00 1.74903846e+00
2.07508713e-01 -1.40957165e+00 6.65847259e-03 6.23536929e-02
1.06724250e+00 -7.29973167e-02 9.80116278e-02 1.61838874e-01
3.08426350e-01 -1.02644360e+00 7.02615380e-01 6.53647900e-01
1.03596485e+00 -6.81250036e-01 5.70535004e-01 -1.14404231e-01
-1.44263065e+00 -1.36151135e-01 -4.89404410e-01 1.72420815e-01
5.35365999e-01 6.02427781e-01 4.38640192e-02 4.66753364e-01
1.14419758e+00 1.33464265e+00 -4.09603387e-01 8.57906520e-01
2.02430785e-01 -2.14318037e-01 1.34518906e-01 7.99516857e-01
-1.14026204e-01 -4.79874134e-01 5.10055006e-01 1.03406727e+00
5.26025653e-01 3.73877972e-01 2.54464835e-01 6.18713260e-01
-1.91665873e-01 -2.37384200e-01 -7.56016791e-01 2.93177009e-01
3.56562495e-01 1.14384294e+00 -3.21737587e-01 -4.28011328e-01
-6.60317659e-01 1.31312251e+00 1.20816574e-01 4.39829886e-01
-1.11403823e+00 -1.76105812e-01 8.47261548e-01 9.74578857e-02
1.68977186e-01 -2.57655919e-01 2.34725550e-02 -1.56526351e+00
-1.53764471e-01 -1.27084780e+00 3.87192667e-01 -1.02735198e+00
-1.23980761e+00 5.81638098e-01 -1.51902422e-01 -1.85010731e+00
-3.00801098e-01 -1.44805619e-02 -3.47449750e-01 6.69301212e-01
-1.53780758e+00 -6.21033430e-01 -8.66459131e-01 8.25898290e-01
8.56490493e-01 -1.34172887e-01 2.20614150e-01 8.20653021e-01
-4.01466012e-01 3.85678619e-01 8.43628794e-02 1.66757047e-01
9.79167163e-01 -7.50056207e-01 3.03765148e-01 1.05769324e+00
-1.15888156e-01 3.38785917e-01 7.07243502e-01 -6.18742168e-01
-1.62301672e+00 -1.16374302e+00 1.95691586e-01 -1.66185737e-01
3.93372774e-01 1.99579652e-02 -1.03427064e+00 6.24116600e-01
2.27539867e-01 4.27570879e-01 2.53741950e-01 -7.62972295e-01
-2.23833665e-01 -4.17213082e-01 -1.04229867e+00 6.90102160e-01
1.23620093e+00 -5.96731961e-01 -3.40892881e-01 3.20499353e-02
9.64542329e-01 -7.67423689e-01 -1.19706750e+00 5.15618265e-01
6.91355884e-01 -1.43982506e+00 1.45814800e+00 2.39861518e-01
8.89782667e-01 -6.67842031e-01 -5.08477509e-01 -1.12436748e+00
-2.86201566e-01 -5.79208851e-01 -2.83899635e-01 1.02895939e+00
-3.20815772e-01 -2.54144311e-01 4.01811838e-01 3.89789462e-01
-5.00554889e-02 -6.83945835e-01 -7.01399386e-01 -3.87076110e-01
-4.21922445e-01 -1.48490936e-01 6.43624842e-01 9.40372229e-01
-4.12669867e-01 -1.09310849e-02 -9.76738393e-01 -1.57340430e-02
9.50945556e-01 4.82977331e-01 8.76521289e-01 -5.20360172e-01
-3.79929900e-01 -1.97803155e-01 -4.60259587e-01 -1.29810822e+00
6.38776869e-02 -4.07857955e-01 2.39899427e-01 -1.31338501e+00
4.35474873e-01 -3.08755279e-01 -1.06967300e-01 -1.11089081e-01
-2.63513654e-01 7.36228108e-01 3.43882143e-01 5.54803431e-01
-7.76404858e-01 6.37190878e-01 1.76661193e+00 -2.71222424e-02
-6.72068521e-02 -6.02778733e-01 -2.69510120e-01 6.19551897e-01
3.42628658e-01 -9.19437706e-02 -1.94269955e-01 -6.01930559e-01
-1.65535465e-01 6.29326105e-01 3.55580270e-01 -1.14467788e+00
5.76775782e-02 -2.86890179e-01 9.66853857e-01 -7.16968298e-01
3.73409063e-01 -8.95241082e-01 7.26281881e-01 4.86284286e-01
-2.79230833e-01 2.68685102e-01 -2.41123661e-01 5.22118270e-01
-5.21317720e-01 2.97870725e-01 1.25461626e+00 -2.29302377e-01
-1.14864719e+00 8.66996348e-01 1.12704955e-01 3.03828530e-03
1.03902352e+00 -4.00250345e-01 -4.59320992e-01 -4.52631176e-01
-3.02353889e-01 -3.34286876e-02 9.76134002e-01 5.77286184e-01
1.02179658e+00 -1.50738299e+00 -7.36969948e-01 4.95772123e-01
-2.23901093e-01 1.34359896e-02 8.22156727e-01 7.48451948e-01
-9.84759808e-01 -4.25592996e-02 -8.09353948e-01 -6.46325231e-01
-1.20146346e+00 8.04585278e-01 4.48737979e-01 -1.90690666e-01
-1.13976276e+00 3.85848135e-01 4.92878884e-01 1.11436285e-01
1.77158788e-01 -3.39112394e-02 -8.40600729e-02 -3.12197030e-01
8.98217678e-01 4.96510535e-01 -3.57724518e-01 -1.04183459e+00
-1.53678268e-01 9.46468830e-01 -9.45341513e-02 -1.06719434e-02
1.24094129e+00 -7.87242115e-01 -1.25615239e-01 1.73553124e-01
1.63122678e+00 -1.67729914e-01 -1.74344134e+00 -2.25052670e-01
-8.02737117e-01 -1.24739861e+00 3.80219780e-02 -1.16232790e-01
-1.55791545e+00 7.23786354e-01 8.27448428e-01 -1.56932071e-01
1.47558773e+00 -3.56839567e-01 1.27450120e+00 1.69568248e-02
3.66359055e-01 -8.87552321e-01 3.05904359e-01 3.36100072e-01
8.38426530e-01 -1.39121354e+00 3.41876209e-01 -2.65562028e-01
-5.92066467e-01 1.07599568e+00 7.81237245e-01 -3.17534477e-01
4.22150224e-01 2.83074468e-01 6.84210211e-02 1.66729778e-01
-6.05104983e-01 7.14555308e-02 6.28677085e-02 6.05903804e-01
3.73813391e-01 -3.63287568e-01 -3.42261977e-02 5.16341627e-02
-1.83561966e-01 7.22736940e-02 6.36342466e-01 6.46305382e-01
-3.02212924e-01 -7.03387737e-01 -5.51871598e-01 2.69669861e-01
-4.19414043e-01 -7.91131631e-02 3.37258935e-01 6.09112263e-01
5.80633208e-02 7.26126313e-01 1.61642343e-01 -6.05052292e-01
1.83128238e-01 -6.80245519e-01 6.81882143e-01 -1.86146377e-03
-1.26318067e-01 2.68763930e-01 -2.31615514e-01 -1.07468390e+00
-6.49159610e-01 -3.92867774e-01 -1.09980035e+00 -8.14237177e-01
2.36267060e-01 -3.50147039e-01 1.93900704e-01 7.46964574e-01
5.15444987e-02 5.32598913e-01 9.69088554e-01 -1.36273229e+00
-1.59364864e-01 -6.49666369e-01 -7.45804846e-01 8.01594675e-01
7.30932951e-01 -5.82788110e-01 -3.67427528e-01 5.05391240e-01] | [10.912384033203125, -2.183711528778076] |
9a81e68f-7377-4098-882e-3eb33c6c9413 | iitp-at-semeval-2017-task-8-a-supervised | null | null | https://aclanthology.org/S17-2087 | https://aclanthology.org/S17-2087.pdf | IITP at SemEval-2017 Task 8 : A Supervised Approach for Rumour Evaluation | This paper describes our system participation in the SemEval-2017 Task 8 {`}RumourEval: Determining rumour veracity and support for rumours{'}. The objective of this task was to predict the stance and veracity of the underlying rumour. We propose a supervised classification approach employing several lexical, content and twitter specific features for learning. Evaluation shows promising results for both the problems. | ['Md. Shad Akhtar', 'Asif Ekbal', 'Pushpak Bhattacharyya', 'Vikram Singh', 'Sunny Narayan'] | 2017-08-01 | null | null | null | semeval-2017-8 | ['rumour-detection'] | ['natural-language-processing'] | [-5.01961946e-01 3.55531871e-01 -6.80394113e-01 -2.55697995e-01
-3.97641450e-01 -2.12774619e-01 1.22753310e+00 5.40414453e-01
-2.20420867e-01 1.26301086e+00 7.35562205e-01 -2.48334453e-01
2.64283270e-01 -4.84523088e-01 -4.64875966e-01 -2.16974914e-01
-1.22392982e-01 4.84720528e-01 3.14084768e-01 -8.25207889e-01
1.18076432e+00 1.34679386e-02 -1.38597572e+00 9.62749362e-01
4.29206580e-01 1.04443324e+00 -5.55573106e-01 7.53791690e-01
-1.03378564e-01 2.26966047e+00 -1.00093126e+00 -4.27431077e-01
-3.75223130e-01 -5.13399124e-01 -1.30938458e+00 5.03572561e-02
4.44296449e-01 -3.50001603e-02 -3.95018935e-01 9.09801185e-01
1.73836410e-01 -6.24966547e-02 8.08610260e-01 -1.33113563e+00
-7.02530563e-01 1.17634475e+00 -4.73849118e-01 9.85461950e-01
6.81926847e-01 -2.06091940e-01 1.20893979e+00 -1.15295410e+00
8.43517900e-01 8.84696126e-01 1.12536716e+00 2.22279072e-01
-8.31516206e-01 -6.82988286e-01 -7.09819496e-01 6.89881980e-01
-1.10273767e+00 -7.28753865e-01 6.14436150e-01 -8.79624963e-01
9.44995284e-01 4.09372121e-01 3.41133654e-01 1.60357845e+00
4.63609576e-01 8.91960025e-01 1.65631342e+00 8.14030170e-02
1.03444539e-01 8.75351846e-01 4.71048146e-01 6.40773833e-01
1.60306796e-01 -2.58639216e-01 -1.04251206e+00 -6.87747717e-01
1.99944582e-02 -5.30035555e-01 -2.65044719e-01 6.74000263e-01
-8.00791740e-01 1.52341127e+00 3.96456808e-01 2.34551549e-01
-7.14270473e-01 -4.15269256e-01 1.00290453e+00 8.28025460e-01
1.34936464e+00 6.63052917e-01 -1.12677552e-01 -2.06890896e-01
-1.47920513e+00 6.14857852e-01 1.66067111e+00 3.75850618e-01
3.60994846e-01 1.75853297e-01 2.40785982e-02 1.05049169e+00
-8.58048722e-03 9.92436931e-02 7.07586467e-01 -6.83076382e-01
2.88876265e-01 1.56517193e-01 5.58908045e-01 -1.30298471e+00
-7.62391925e-01 -4.78282779e-01 -5.83370090e-01 -1.92230921e-02
1.79317370e-02 -2.79939562e-01 -3.85854483e-01 7.43217289e-01
-7.08180070e-02 4.09829438e-01 3.51261459e-02 8.10911119e-01
1.53274107e+00 7.85908699e-01 -7.67509500e-03 -7.41508603e-01
1.05207050e+00 -1.01436257e+00 -9.87736881e-01 -5.56529984e-02
6.33551836e-01 -1.15837884e+00 4.77257818e-01 5.30598342e-01
-1.23248780e+00 2.07289323e-01 -1.07861114e+00 2.14134216e-01
-3.42904955e-01 -3.05925816e-01 2.23774433e-01 4.26252425e-01
-8.44301641e-01 8.41916442e-01 7.06195980e-02 -5.22307120e-02
5.01096666e-01 -3.24471831e-01 -4.39525060e-02 1.52535543e-01
-1.63641107e+00 1.26378715e+00 4.06654626e-01 -5.22414923e-01
-7.83463597e-01 -4.67298299e-01 -4.80545312e-01 -4.54522163e-01
-8.76386911e-02 -1.66110545e-01 1.48478031e+00 -8.91747177e-01
-1.31264877e+00 1.38036120e+00 4.65918370e-02 -1.02038181e+00
9.61480379e-01 -9.04678032e-02 -8.67351115e-01 -1.27802789e-01
3.04647475e-01 -3.70868236e-01 9.88184333e-01 -9.96761978e-01
-5.12854993e-01 -1.39018193e-01 -5.03107250e-01 -2.09247842e-01
-1.60710171e-01 6.74381018e-01 7.37252414e-01 -5.89679837e-01
-7.19464943e-02 -5.75012684e-01 3.92085165e-01 -1.13002467e+00
-6.93475962e-01 -6.30776286e-01 6.97705150e-01 -1.06577158e+00
1.57124710e+00 -1.51233125e+00 -3.22303414e-01 1.10750645e-01
4.51362193e-01 2.45734692e-01 4.74373192e-01 7.94422746e-01
1.80319861e-01 2.28095740e-01 3.39262217e-01 -2.08633453e-01
-2.14466125e-01 -1.12160929e-01 -9.75425184e-01 1.03011417e+00
-2.82453954e-01 5.41506708e-01 -8.12748790e-01 -4.19266701e-01
-4.61323559e-01 6.25757352e-02 -2.26225689e-01 2.96411991e-01
-3.54391754e-01 1.55574411e-01 -2.76240289e-01 5.37499845e-01
3.71916920e-01 -4.70193088e-01 -2.09739909e-01 3.08273166e-01
-3.34902436e-01 1.06192374e+00 -2.43617386e-01 4.20794398e-01
-1.89947397e-01 9.15174425e-01 5.04758507e-02 -5.82222402e-01
1.22115874e+00 6.46353304e-01 3.35826635e-01 -4.89260763e-01
4.98189420e-01 7.67193258e-01 -5.22651970e-01 -7.67873943e-01
8.62447679e-01 -6.24491453e-01 -2.31368884e-01 8.48877668e-01
-3.29413831e-01 -1.64186116e-03 -3.38357598e-01 4.79477882e-01
9.07479942e-01 -4.63612825e-01 8.82762790e-01 -4.59903479e-01
6.62149072e-01 5.04908025e-01 9.54688266e-02 8.04476023e-01
-3.12455088e-01 3.26736897e-01 8.27060640e-01 -6.41557753e-01
-1.08857369e+00 -3.86510253e-01 -2.64227659e-01 1.48938310e+00
-3.33799511e-01 -3.13837558e-01 -2.41094932e-01 -6.38421118e-01
2.98730850e-01 1.40157390e+00 -1.00996363e+00 3.51003081e-01
-3.40422183e-01 -8.21826875e-01 9.45748389e-01 -1.07468516e-01
2.20244557e-01 -1.19540310e+00 -1.59130916e-02 4.26870584e-01
-5.15129387e-01 -1.15283597e+00 2.64093615e-02 -1.74828187e-01
-5.96218169e-01 -9.81322587e-01 -2.46670127e-01 -5.28064430e-01
-2.89685279e-01 1.17372200e-01 1.23210478e+00 4.33418363e-01
2.61262387e-01 -2.93928474e-01 -3.39132607e-01 -5.32529116e-01
-9.74219799e-01 3.77928801e-02 2.26151854e-01 -2.01221153e-01
3.32824230e-01 -7.57188261e-01 -2.12424889e-01 3.84183288e-01
-4.16270703e-01 -2.32800290e-01 -7.06352890e-02 7.32014060e-01
-1.91083789e-01 -2.58822411e-01 1.34699512e+00 -1.48086679e+00
1.33350050e+00 -1.69948518e+00 2.75013506e-01 -4.66370761e-01
-9.54657793e-01 -6.61675572e-01 4.82535869e-01 2.34389100e-02
-5.84254742e-01 -1.03583360e+00 -3.66384566e-01 -6.07804283e-02
1.50531247e-01 8.57049823e-01 9.17357862e-01 2.02832252e-01
1.36961889e+00 3.82222444e-01 6.02835417e-02 -5.64623594e-01
1.95679627e-02 1.40262985e+00 3.41075778e-01 1.11363560e-01
1.85031623e-01 3.08030605e-01 -3.98272514e-01 -1.00738144e+00
-1.62569833e+00 -1.01642334e+00 -9.06814113e-02 -3.14317465e-01
1.72762185e-01 -1.07573462e+00 -6.55262113e-01 4.56592977e-01
-1.34234333e+00 1.04559503e-01 1.74792364e-01 1.30979359e-01
-6.59501672e-01 -3.81434709e-02 -1.29864717e+00 -8.95365655e-01
-7.44556308e-01 -1.93942532e-01 6.61358908e-02 -2.85826206e-01
-6.99197948e-01 -1.19862056e+00 6.39602184e-01 8.00907195e-01
5.74324369e-01 4.28591996e-01 3.52247328e-01 -1.70277190e+00
3.38740796e-01 -5.31400979e-01 -4.81285185e-01 1.41782209e-01
-3.39422554e-01 -2.53045470e-01 -9.67909217e-01 -2.08395407e-01
9.62198600e-02 -1.07879257e+00 1.19169307e+00 -1.59249470e-01
6.54744685e-01 -1.42644358e+00 1.66241303e-02 -1.39544725e-01
1.07463753e+00 -9.54986930e-01 3.52529883e-01 8.66391242e-01
3.38674098e-01 6.77697241e-01 5.64210236e-01 1.24384284e+00
5.37880123e-01 4.10153747e-01 4.81547087e-01 6.92378700e-01
-7.95914307e-02 -3.24621260e-01 8.40917230e-01 1.01368010e+00
-2.40045100e-01 -2.83818901e-01 -1.01556528e+00 6.25761151e-01
-1.82774734e+00 -1.57938766e+00 -8.96019638e-01 1.45573485e+00
1.14361703e+00 3.37566018e-01 6.25122428e-01 5.06318025e-02
4.71827060e-01 7.81220734e-01 5.02296127e-02 -1.07672787e+00
-2.65224904e-01 -4.04493749e-01 4.14974988e-01 7.50565112e-01
-8.89324427e-01 8.11645687e-01 7.07651424e+00 4.37176794e-01
-1.24881697e+00 7.35784888e-01 2.90032297e-01 -2.99223512e-01
-2.89204955e-01 -4.61997658e-01 -8.62622321e-01 3.99811059e-01
1.42814326e+00 -5.05493343e-01 1.51865616e-01 7.61659443e-01
5.13460338e-01 -8.92411172e-02 -5.70661902e-01 4.09373909e-01
8.64053726e-01 -1.83273375e+00 -3.08576435e-01 -6.17022701e-02
1.03989637e+00 9.48341072e-01 2.36141887e-02 6.63203001e-01
3.88512522e-01 -1.13632333e+00 1.04323471e+00 4.34422791e-01
2.29240373e-01 -5.31944633e-01 1.03160131e+00 9.93473351e-01
1.38279080e-01 -1.86055545e-02 -3.55586320e-01 -3.84875208e-01
1.64632097e-01 8.44943404e-01 -1.66404402e+00 -1.02821171e-01
5.00975132e-01 1.02266645e+00 -4.73921925e-01 8.38355660e-01
-4.61991966e-01 1.32694256e+00 2.63977826e-01 -2.76805192e-01
4.88723814e-01 5.31990767e-01 1.13159049e+00 1.84957814e+00
-1.10876970e-01 -4.72178170e-03 2.72145987e-01 5.60861588e-01
-6.35164857e-01 5.84535182e-01 -5.29482782e-01 -8.44775513e-02
3.35171342e-01 8.70640814e-01 -1.26751378e-01 -4.55587804e-01
-5.30486405e-02 8.51116478e-01 4.70794320e-01 -2.29477242e-01
-7.89775908e-01 1.83362216e-01 1.93071797e-01 7.72652149e-01
1.60896838e-01 2.24415347e-01 -6.91180587e-01 -1.09730387e+00
-6.76275253e-01 -8.74947429e-01 7.06323802e-01 -5.24185002e-01
-1.68923378e+00 7.41962314e-01 -3.66942018e-01 -7.45610654e-01
-3.62317741e-01 -1.22934058e-01 -1.01868343e+00 8.23143542e-01
-1.88773656e+00 -1.08434772e+00 -1.60188496e-01 5.49494863e-01
6.67074203e-01 -6.70181453e-01 5.54860055e-01 -1.47738859e-01
-3.22057039e-01 2.21376508e-01 1.15763247e-01 -1.74525514e-01
9.95662630e-01 -1.09339583e+00 2.46717725e-02 -7.90724754e-02
-2.71040231e-01 9.02881622e-02 1.73792720e+00 -8.66328537e-01
-6.47687137e-01 -1.08316720e+00 1.75419748e+00 -7.67043173e-01
1.68609726e+00 3.29514146e-01 -1.24473548e+00 7.58736312e-01
5.29476762e-01 -1.44888088e-01 9.27422225e-01 4.09768760e-01
-8.40306520e-01 7.21784711e-01 -1.30618966e+00 -1.07752442e-01
2.39649609e-01 -6.23156130e-01 -1.04737616e+00 9.44048703e-01
2.31669843e-01 -2.31600970e-01 -1.00652218e+00 -1.70299232e-01
3.84161949e-01 -1.21845829e+00 4.67247099e-01 -1.26933455e+00
1.35438669e+00 3.18212181e-01 -1.28649399e-01 -1.44327235e+00
-3.51217210e-01 -6.60380840e-01 -5.72103798e-01 9.27558243e-01
6.40069425e-01 -6.21111095e-01 7.62315094e-01 -1.69655696e-01
6.82872683e-02 -6.15097761e-01 -7.99685299e-01 -5.97856343e-01
4.13297057e-01 -2.77979732e-01 -5.63375093e-02 1.62973630e+00
8.35765362e-01 7.56411016e-01 -1.08968794e+00 -9.98785123e-02
6.47913039e-01 2.42139637e-01 4.89278287e-01 -1.30214489e+00
-4.45766523e-02 -6.79295361e-01 -3.69464844e-01 -3.86477202e-01
4.66995865e-01 -1.19553947e+00 -5.08878194e-02 -1.18053603e+00
3.75440717e-01 -2.07905069e-01 -3.61444235e-01 1.63671285e-01
1.12247340e-01 6.75311446e-01 -2.23592371e-01 1.00762558e+00
-7.27657557e-01 3.24039638e-01 8.25007021e-01 5.68410708e-03
-4.61719781e-02 6.12357676e-01 -5.77126563e-01 8.13249409e-01
1.17248642e+00 -6.82557344e-01 2.68876255e-01 4.25314516e-01
8.60278964e-01 4.31295246e-01 5.23693621e-01 -3.95010591e-01
2.30122149e-01 -3.44668269e-01 -2.95141637e-01 -7.16355920e-01
1.79414272e-01 1.00191154e-01 -4.00514901e-01 3.76260251e-01
-6.24800444e-01 -1.41795659e-02 -6.98897988e-02 7.15630293e-01
-3.35620373e-01 -3.93145919e-01 1.08227479e+00 -2.23760217e-01
-4.94642645e-01 -6.08770140e-02 -6.58795178e-01 4.40337807e-01
7.69307673e-01 2.81378359e-01 -8.98935735e-01 -8.96514535e-01
-8.95157337e-01 -1.26670450e-01 2.85517007e-01 5.20661116e-01
6.57176316e-01 -1.05774486e+00 -1.85126841e+00 -4.26910222e-01
3.34821939e-01 -1.26641345e+00 -1.98876977e-01 1.59532940e+00
-4.36840236e-01 3.93479615e-01 -2.67673641e-01 -2.91442964e-02
-1.25903952e+00 3.24787766e-01 3.03498447e-01 -1.35453701e-01
-6.24774933e-01 8.53233814e-01 -9.38318610e-01 -6.14014417e-02
-1.74469218e-01 8.42450500e-01 -9.67891097e-01 6.55298412e-01
1.08752346e+00 9.63232398e-01 2.70398438e-01 -1.17780304e+00
-6.98124105e-03 -7.91497290e-01 -4.64645475e-01 -3.70219462e-02
1.60907805e+00 -3.25163484e-01 -6.70251548e-01 1.08063853e+00
1.24381483e+00 2.82966733e-01 -2.57128000e-01 -7.11652040e-01
5.92080355e-01 -4.97463942e-01 5.97131014e-01 -1.10491240e+00
-4.39428180e-01 4.38283682e-01 -4.19028699e-01 9.09889042e-01
-7.06850290e-02 1.39643788e-01 9.77771819e-01 8.59485939e-02
1.33643627e-01 -1.16908586e+00 6.95373937e-02 1.10302746e+00
1.46396935e+00 -1.43818069e+00 6.08074307e-01 -2.62006909e-01
-1.04138672e+00 1.22877288e+00 2.08990276e-01 -2.82073557e-01
7.89193809e-01 -2.86033928e-01 1.03784986e-01 -7.76878059e-01
-1.18776369e+00 4.53955919e-01 3.39707196e-01 -5.69490306e-02
7.20973134e-01 3.24407637e-01 -8.92857969e-01 6.63141370e-01
-6.49264455e-01 -5.14027523e-03 1.34631002e+00 5.72548985e-01
-1.18738222e+00 -1.21742405e-01 -3.47857922e-01 5.88938534e-01
-1.14744127e+00 -3.62367183e-02 -8.19576919e-01 4.72676992e-01
-1.81414530e-01 1.02296638e+00 -3.88552219e-01 -6.56949222e-01
-2.69093402e-02 2.03364030e-01 -3.73051539e-02 -6.96185946e-01
-1.07620442e+00 -3.55162770e-01 1.09102964e+00 -4.53356832e-01
-3.37113142e-01 -1.03925037e+00 -9.75236595e-01 -1.17864430e+00
-2.15655476e-01 5.14481246e-01 5.97257197e-01 1.02733886e+00
-1.98848009e-01 8.17025453e-02 1.20234823e+00 -3.72284174e-01
-1.18943477e+00 -1.40572226e+00 -9.70858395e-01 5.37776470e-01
6.22803271e-01 -3.08110118e-01 -7.60215104e-01 -1.85736090e-01] | [8.228715896606445, 10.112257957458496] |
4a5d03bf-bbf0-4362-a1e4-40b850d1f27c | use-of-affective-visual-information-for | 2107.03783 | null | https://arxiv.org/abs/2107.03783v1 | https://arxiv.org/pdf/2107.03783v1.pdf | Use of Affective Visual Information for Summarization of Human-Centric Videos | Increasing volume of user-generated human-centric video content and their applications, such as video retrieval and browsing, require compact representations that are addressed by the video summarization literature. Current supervised studies formulate video summarization as a sequence-to-sequence learning problem and the existing solutions often neglect the surge of human-centric view, which inherently contains affective content. In this study, we investigate the affective-information enriched supervised video summarization task for human-centric videos. First, we train a visual input-driven state-of-the-art continuous emotion recognition model (CER-NET) on the RECOLA dataset to estimate emotional attributes. Then, we integrate the estimated emotional attributes and the high-level representations from the CER-NET with the visual information to define the proposed affective video summarization architectures (AVSUM). In addition, we investigate the use of attention to improve the AVSUM architectures and propose two new architectures based on temporal attention (TA-AVSUM) and spatial attention (SA-AVSUM). We conduct video summarization experiments on the TvSum database. The proposed AVSUM-GRU architecture with an early fusion of high level GRU embeddings and the temporal attention based TA-AVSUM architecture attain competitive video summarization performances by bringing strong performance improvements for the human-centric videos compared to the state-of-the-art in terms of F-score and self-defined face recall metrics. | ['Engin Erzin', 'Berkay Köprü'] | 2021-07-08 | null | null | null | null | ['supervised-video-summarization'] | ['computer-vision'] | [ 1.05866730e-01 -1.31694376e-01 -1.43115535e-01 -2.18253240e-01
-7.96506047e-01 -1.17471732e-01 5.75520337e-01 2.46626347e-01
-4.02345121e-01 5.07541776e-01 8.66160452e-01 3.94553095e-01
9.17645320e-02 -1.78211927e-01 -6.88873708e-01 -6.27358973e-01
-2.53759455e-02 -5.88131361e-02 -3.06174532e-03 -1.01174213e-01
3.18482816e-01 4.71863570e-03 -1.85898638e+00 5.85311592e-01
9.66977656e-01 1.31404757e+00 -2.90598609e-02 1.08641732e+00
-2.10915342e-01 1.12455535e+00 -5.38459539e-01 -4.33855653e-01
-2.16203213e-01 -7.74624527e-01 -4.72436517e-01 1.39995202e-01
5.86601794e-01 -3.28006297e-01 -5.64243913e-01 9.58867729e-01
7.46507227e-01 6.45065069e-01 8.89230728e-01 -1.27123022e+00
-8.90234351e-01 1.72822535e-01 -7.99988866e-01 2.46822670e-01
6.14621699e-01 -8.40198062e-03 9.60032821e-01 -1.07158744e+00
8.54860723e-01 1.10321212e+00 4.37357903e-01 5.26951134e-01
-7.30907619e-01 -2.55871624e-01 3.42770308e-01 7.52229273e-01
-1.21887267e+00 -4.85298097e-01 9.53442097e-01 -3.91424268e-01
1.19825578e+00 5.66461623e-01 8.48369658e-01 1.29199386e+00
2.42086232e-01 1.19377446e+00 4.69576478e-01 -8.09233785e-02
3.15634042e-01 3.63367982e-02 3.89928073e-01 5.67648947e-01
4.37451489e-02 -4.95748967e-01 -6.06493711e-01 1.69378117e-01
2.20744714e-01 1.86368883e-01 -2.54422188e-01 -2.99545556e-01
-9.73568559e-01 6.90148652e-01 2.66195029e-01 2.81229973e-01
-8.90600562e-01 9.68177766e-02 1.18657064e+00 1.13730289e-01
7.86361694e-01 3.09974611e-01 1.13736182e-01 -2.04621464e-01
-1.35596490e+00 1.90141737e-01 4.42737937e-01 9.30990338e-01
4.42668438e-01 5.33200562e-01 -7.11569250e-01 9.55179811e-01
5.43426629e-03 2.93658108e-01 8.61318767e-01 -9.70304310e-01
9.58520919e-02 5.85368097e-01 -1.95725799e-01 -1.23859179e+00
-9.91521478e-02 -2.93791473e-01 -9.56766427e-01 -2.61004448e-01
-7.86102176e-01 -1.10834725e-01 -8.60879838e-01 1.61147976e+00
4.13104286e-03 4.04844314e-01 4.24460471e-01 9.31613326e-01
1.49477196e+00 1.31519747e+00 3.40996265e-01 -8.06837499e-01
1.20608330e+00 -1.51019847e+00 -1.14618313e+00 8.63091275e-02
4.24667388e-01 -3.69853288e-01 9.56891954e-01 2.95749366e-01
-1.16575074e+00 -7.71873713e-01 -1.25490022e+00 -1.23169847e-01
-2.89372236e-01 2.09012091e-01 1.23272374e-01 2.24877656e-01
-1.34021246e+00 3.64668906e-01 -4.65649754e-01 -7.93074965e-01
4.50344980e-01 1.66867405e-01 -5.95295310e-01 9.47074071e-02
-1.07781792e+00 6.85683131e-01 6.66021109e-01 -3.79600413e-02
-8.34959388e-01 -4.13392842e-01 -1.07596159e+00 2.70169973e-01
4.24141407e-01 -8.73487592e-01 9.69562471e-01 -1.58470511e+00
-1.42375684e+00 6.27959192e-01 -1.92262247e-01 -5.18806100e-01
2.33318269e-01 -5.94721317e-01 -4.85628575e-01 7.56515741e-01
-1.24311842e-01 8.10333133e-01 1.03459728e+00 -1.29180932e+00
-4.43916202e-01 -2.44565412e-01 -2.98526883e-01 6.24249279e-01
-7.92739153e-01 7.29974285e-02 -5.33459723e-01 -8.77703726e-01
-6.09840631e-01 -7.10269034e-01 5.06301485e-02 -5.37917197e-01
-1.46971256e-01 -2.88512141e-01 1.23120785e+00 -1.07601988e+00
1.69607592e+00 -2.19659591e+00 6.43836915e-01 -2.65337706e-01
1.85189992e-01 6.56710863e-01 -4.05668378e-01 4.83509004e-01
-2.34904200e-01 6.36759996e-02 -1.00820974e-01 -5.26329339e-01
-8.27633962e-02 -1.31671011e-01 -5.54263368e-02 1.18710637e-01
2.28351593e-01 1.06063533e+00 -8.48522484e-01 -9.03960109e-01
3.64412814e-01 4.95442212e-01 -7.03668892e-01 5.69213569e-01
-1.70810312e-01 1.13768339e-01 -3.44664603e-01 5.49378932e-01
4.56195354e-01 1.80470571e-01 1.52565911e-02 -6.52019262e-01
-9.52922106e-02 -7.04425216e-01 -5.99790096e-01 1.84510612e+00
-1.83865190e-01 6.67412460e-01 -2.38101721e-01 -1.05820382e+00
7.66960263e-01 4.59063917e-01 7.67815650e-01 -6.29832149e-01
4.28674817e-01 -1.43936485e-01 -4.60607648e-01 -1.01690364e+00
1.10459101e+00 -5.23212105e-02 -2.34711319e-01 8.54201615e-02
4.97831225e-01 3.25233579e-01 2.22246245e-01 6.35018051e-01
8.87760520e-01 3.02539438e-01 5.76928258e-01 4.40715812e-02
8.28524888e-01 -2.53133714e-01 5.55187821e-01 3.59060824e-01
-3.21151942e-01 6.69248402e-01 5.89164078e-01 -1.40809104e-01
-1.00624847e+00 -7.53456116e-01 5.19754946e-01 1.39375484e+00
1.93886980e-01 -7.40850449e-01 -9.86684084e-01 -7.41178453e-01
-3.06072503e-01 7.92129397e-01 -6.98654294e-01 -5.25046468e-01
-3.39079499e-01 -4.50437993e-01 3.34464282e-01 6.90407336e-01
6.20657086e-01 -1.22558296e+00 -7.49870062e-01 -9.99618322e-03
-5.00585258e-01 -1.05312908e+00 -5.98247766e-01 -4.37609673e-01
-4.44682747e-01 -7.27178156e-01 -1.04134488e+00 -7.19328701e-01
3.58158171e-01 9.48950350e-02 7.95358539e-01 -3.17774802e-01
-9.80936363e-02 9.18557703e-01 -8.51054251e-01 -2.21579537e-01
-9.61880088e-02 -9.21185613e-02 1.76877320e-01 5.44424772e-01
3.42128634e-01 -4.92522776e-01 -6.45584404e-01 -3.55806708e-01
-1.02706540e+00 2.40700603e-01 7.58030534e-01 6.98409379e-01
6.26566648e-01 -4.02284950e-01 1.06566274e+00 -6.05320752e-01
9.05850530e-01 -7.17783630e-01 3.88208359e-01 4.12194699e-01
-2.57038265e-01 -1.37618408e-01 5.15436172e-01 -4.11084384e-01
-1.32869589e+00 -1.48317710e-01 -1.14400603e-01 -1.09531617e+00
3.28569114e-02 7.04485238e-01 -2.05067843e-01 3.31827492e-01
2.54109591e-01 4.27877575e-01 -8.67210999e-02 -5.27575500e-02
5.65788507e-01 9.24609125e-01 7.82247126e-01 -3.28773826e-01
2.99377926e-02 4.73133139e-02 -3.84996027e-01 -1.31778383e+00
-7.03131557e-01 -7.16530502e-01 -2.25227386e-01 -8.25595617e-01
1.42288375e+00 -9.59992826e-01 -5.65399587e-01 4.14003968e-01
-1.21348608e+00 2.66406059e-01 -3.84025991e-01 3.58834952e-01
-8.80527437e-01 7.98074961e-01 -5.05171776e-01 -9.71799672e-01
-9.05211270e-01 -9.21428442e-01 1.09417188e+00 5.04999161e-01
-3.98135215e-01 -6.25886798e-01 2.02411383e-01 3.35554779e-01
3.42507243e-01 5.34247339e-01 8.11097026e-01 -9.02642012e-01
-1.00777529e-01 -1.49359494e-01 -2.26359800e-01 7.11682379e-01
-3.14841598e-01 1.47545874e-01 -8.85865927e-01 -2.52728611e-01
-7.63402283e-02 -5.90679526e-01 1.35155487e+00 5.94796896e-01
1.18595088e+00 -4.99814898e-01 5.42521887e-02 3.79752845e-01
1.28819168e+00 2.73551404e-01 9.34191346e-01 -6.16596565e-02
7.82920718e-01 4.94576871e-01 7.58170366e-01 7.20864892e-01
3.54712069e-01 3.85823756e-01 3.91365737e-01 1.41384229e-01
-3.27169374e-02 -2.23838434e-01 6.73293829e-01 1.43155456e+00
-4.48256195e-01 -5.65199077e-01 -3.31744790e-01 5.85475981e-01
-2.42475748e+00 -1.28879511e+00 1.68279380e-01 1.76904345e+00
3.20403516e-01 -3.15621614e-01 2.26652384e-01 -1.04795694e-01
8.04856002e-01 6.28174543e-01 -5.06374359e-01 -8.80105317e-01
-2.06647977e-01 -1.53707892e-01 -1.02775119e-01 1.18954673e-01
-1.22581327e+00 9.29267049e-01 5.09520483e+00 1.02758026e+00
-9.54574883e-01 1.91642985e-01 6.77560925e-01 -5.04807711e-01
-1.96126148e-01 -5.51830649e-01 -1.07449412e-01 5.60442984e-01
1.18940473e+00 -4.13005590e-01 -5.55626079e-02 8.23562801e-01
2.07215592e-01 -1.55939639e-01 -1.05704772e+00 1.45680261e+00
1.08561075e+00 -1.22960055e+00 7.27837861e-01 -3.57544273e-01
8.78396571e-01 -4.49518949e-01 1.21893398e-01 8.07875514e-01
-4.64373589e-01 -7.50297010e-01 6.62930608e-01 9.59700286e-01
9.31592762e-01 -9.66055572e-01 1.10251796e+00 -6.02230849e-03
-1.26578617e+00 -1.68118790e-01 -1.68038726e-01 2.53765315e-01
2.76856244e-01 9.38874185e-02 -1.21760510e-01 1.00042033e+00
5.46549261e-01 1.22506535e+00 -6.70633078e-01 8.08421433e-01
1.91958666e-01 4.31713611e-01 3.36994022e-01 -9.65728462e-02
4.04017150e-01 -2.76000559e-01 8.24539304e-01 1.60014200e+00
3.97841424e-01 3.83058369e-01 4.38152552e-02 1.55368358e-01
-3.46768051e-01 4.88348812e-01 -6.20942593e-01 -4.90400612e-01
6.20465465e-02 1.37148190e+00 -4.52191353e-01 -6.98100269e-01
-3.68003160e-01 1.38136029e+00 1.80172071e-01 4.67282742e-01
-1.07351899e+00 -4.69555795e-01 4.86223757e-01 -1.92378566e-01
3.94732326e-01 1.85969815e-01 3.04780543e-01 -1.37002313e+00
-1.50009319e-01 -8.85787129e-01 5.02018631e-01 -1.11943269e+00
-1.05020833e+00 7.81674922e-01 1.44333303e-01 -1.21522188e+00
-3.94121200e-01 -1.77118018e-01 -7.20017970e-01 4.88648415e-02
-1.10470629e+00 -1.23827648e+00 -5.82806408e-01 5.54253161e-01
1.12666714e+00 -4.26016927e-01 6.34668052e-01 3.36885363e-01
-7.91413605e-01 5.44590473e-01 9.23442394e-02 -2.53410727e-01
8.87616515e-01 -7.91993737e-01 -3.42258543e-01 7.88734674e-01
-1.11454323e-01 1.77189291e-01 8.68645251e-01 -7.14667737e-01
-1.43312371e+00 -1.37385774e+00 5.83549201e-01 -1.49840182e-02
3.38891029e-01 5.71815297e-02 -9.07667518e-01 7.54542708e-01
8.32859933e-01 -2.20108941e-01 8.31266165e-01 -2.62964338e-01
-1.37384146e-01 -3.71447392e-02 -1.00143015e+00 6.03944361e-01
8.79231691e-01 -4.01657224e-01 -8.79150808e-01 -8.55882838e-02
8.71608794e-01 1.12521566e-01 -6.33581400e-01 5.01169205e-01
6.43421888e-01 -9.44803715e-01 7.79583752e-01 -8.57998788e-01
1.00217927e+00 -1.09249771e-01 -3.08814019e-01 -1.45038342e+00
-4.05782014e-01 -3.76971096e-01 -4.88865048e-01 1.47522950e+00
-2.54121333e-01 4.93349358e-02 3.99870664e-01 1.26054764e-01
-4.26054269e-01 -8.87962461e-01 -6.23892903e-01 -3.79745215e-01
-4.98535782e-01 -8.78021121e-02 2.82163657e-02 8.33099604e-01
2.59411961e-01 6.88984752e-01 -8.35297704e-01 -3.88122946e-01
4.25815344e-01 -1.06030539e-01 5.98972797e-01 -8.98130119e-01
1.69314340e-01 -4.73864555e-01 -8.59259069e-01 -5.66895723e-01
5.72608531e-01 -8.75999629e-01 -8.57892558e-02 -1.85310388e+00
7.75291741e-01 7.71395326e-01 -5.73214710e-01 3.53186354e-02
-2.75647163e-01 2.10638985e-01 4.28447604e-01 -1.32051945e-01
-1.56217122e+00 1.25914407e+00 6.65659547e-01 -2.31443018e-01
-9.43216458e-02 -7.25608587e-01 -5.88447332e-01 7.37415969e-01
5.08080781e-01 1.43271424e-02 -6.35883868e-01 -8.73215646e-02
1.11748040e-01 3.25061142e-01 1.93327352e-01 -1.23715687e+00
1.72798738e-01 5.02230376e-02 4.44292188e-01 -9.21560705e-01
5.31809688e-01 -5.96979976e-01 4.72790524e-02 1.21111415e-01
-5.26704371e-01 1.98895782e-01 1.82010666e-01 8.82105887e-01
-6.13566339e-01 -3.72875407e-02 6.53902650e-01 1.07254293e-02
-1.16329622e+00 3.37003469e-01 -6.55322850e-01 3.51458639e-02
1.46213233e+00 -4.44674879e-01 -2.18782455e-01 -7.80896723e-01
-6.94869757e-01 3.58708084e-01 2.76583493e-01 5.98826766e-01
1.14481485e+00 -1.50224841e+00 -9.62807655e-01 -7.34911412e-02
3.41755629e-01 -5.78157067e-01 1.04240310e+00 8.31587791e-01
-4.51886475e-01 3.47419113e-01 -6.90178156e-01 -4.04300064e-01
-1.61672473e+00 8.31164241e-01 -1.51039302e-01 -1.57879427e-01
-4.25651491e-01 6.69961393e-01 4.66380715e-01 2.22886592e-01
1.75783247e-01 1.91636205e-01 -7.36878514e-01 6.45550251e-01
5.00772417e-01 6.79457128e-01 -2.46719569e-01 -9.81713295e-01
-3.48823756e-01 5.05995393e-01 -1.35753453e-01 -8.86124093e-03
1.43538451e+00 -2.32162103e-01 -2.79635862e-02 5.08826077e-01
1.52774894e+00 -3.67261559e-01 -9.25346076e-01 2.98946165e-02
-2.07230255e-01 -5.33182323e-02 2.49049459e-02 -4.59259242e-01
-1.08860993e+00 9.25356090e-01 5.94752610e-01 -2.66484231e-01
1.34534252e+00 -1.36383235e-01 1.04680336e+00 2.12955132e-01
-1.76589265e-01 -1.44877708e+00 5.75414419e-01 3.41673464e-01
1.21689343e+00 -1.06718528e+00 1.28147751e-01 6.17891923e-02
-1.42440498e+00 8.19428861e-01 7.50293076e-01 -1.31032512e-01
1.56570300e-01 -2.62968183e-01 -3.71886969e-01 -4.59195152e-02
-1.06708610e+00 -1.45907611e-01 5.43053210e-01 3.86300921e-01
3.27151924e-01 -4.79410589e-02 -5.82459152e-01 1.15202630e+00
2.33644620e-01 1.03444003e-01 6.21824622e-01 6.90767765e-01
-6.29776001e-01 -3.61345619e-01 -4.10597026e-02 6.24973059e-01
-4.04387444e-01 3.43917008e-03 -5.68767846e-01 3.69943917e-01
-1.18720278e-01 7.92372942e-01 2.21720606e-01 -7.20525920e-01
3.55834186e-01 3.55657876e-01 3.65869999e-01 -3.50824654e-01
-6.81258559e-01 1.47824898e-01 2.46510521e-01 -5.58597028e-01
-7.33285189e-01 -5.45249164e-01 -1.18688047e+00 -1.48501337e-01
1.87989399e-01 3.53249162e-01 3.65670711e-01 7.11927593e-01
6.11782134e-01 9.58407879e-01 5.18333852e-01 -1.16361988e+00
-1.34960845e-01 -1.00924468e+00 -4.07822311e-01 7.06827581e-01
2.66913205e-01 -4.82812762e-01 -1.64388716e-01 2.25445271e-01] | [10.446755409240723, 0.44722074270248413] |
61fb7de2-01d1-4dab-a85e-6bb485ce3562 | interactive-feature-embedding-for-infrared | 2211.04877 | null | https://arxiv.org/abs/2211.04877v1 | https://arxiv.org/pdf/2211.04877v1.pdf | Interactive Feature Embedding for Infrared and Visible Image Fusion | General deep learning-based methods for infrared and visible image fusion rely on the unsupervised mechanism for vital information retention by utilizing elaborately designed loss functions. However, the unsupervised mechanism depends on a well designed loss function, which cannot guarantee that all vital information of source images is sufficiently extracted. In this work, we propose a novel interactive feature embedding in self-supervised learning framework for infrared and visible image fusion, attempting to overcome the issue of vital information degradation. With the help of self-supervised learning framework, hierarchical representations of source images can be efficiently extracted. In particular, interactive feature embedding models are tactfully designed to build a bridge between the self-supervised learning and infrared and visible image fusion learning, achieving vital information retention. Qualitative and quantitative evaluations exhibit that the proposed method performs favorably against state-of-the-art methods. | ['Huchuan Lu', 'Wenda Zhao', 'Fan Zhao'] | 2022-11-09 | null | null | null | null | ['infrared-and-visible-image-fusion'] | ['computer-vision'] | [ 2.41108596e-01 -8.33313689e-02 -2.96346724e-01 -1.46837473e-01
-6.47171140e-01 -7.04369843e-02 4.65568811e-01 1.78604782e-01
-2.77697831e-01 5.66385984e-01 1.96868554e-01 -8.73056278e-02
-3.82803887e-01 -8.19332361e-01 -3.32785636e-01 -1.25563693e+00
3.19187671e-01 -4.65925604e-01 -2.72157520e-01 -2.89935470e-01
2.02084463e-02 4.31228280e-01 -1.80166209e+00 1.44069016e-01
1.13872683e+00 1.21126509e+00 9.20411870e-02 3.89349461e-01
-2.82102317e-01 9.29964781e-01 -2.84433603e-01 -4.92109917e-02
2.75207251e-01 -5.86125612e-01 -5.22189200e-01 1.94177464e-01
2.38612026e-01 -2.08628982e-01 -5.39753377e-01 1.00689065e+00
7.06507862e-01 -4.26091291e-02 7.37382829e-01 -1.11349285e+00
-7.63071597e-01 3.33295465e-01 -5.25259197e-01 3.80491652e-03
3.58792394e-01 2.71103591e-01 9.08086002e-01 -7.89638400e-01
3.41639936e-01 7.71478653e-01 5.97399414e-01 3.74986321e-01
-1.26252282e+00 -6.50990844e-01 -4.57088873e-02 1.61621168e-01
-1.18904746e+00 -5.87175906e-01 1.25224340e+00 -4.75003630e-01
4.79765505e-01 4.19766963e-01 7.15739012e-01 7.40458727e-01
4.87302274e-01 6.44245982e-01 1.17296219e+00 -6.58441544e-01
1.74828976e-01 4.82534200e-01 1.48969382e-01 8.28051746e-01
2.79997200e-01 4.55871314e-01 -7.69704640e-01 -1.02958791e-01
4.12130207e-01 5.22129118e-01 -3.37686270e-01 -3.88743490e-01
-8.91068399e-01 6.74444377e-01 6.72696710e-01 7.83590317e-01
-4.12819773e-01 -3.89759392e-02 6.91998228e-02 5.73270977e-01
6.96609199e-01 5.76234274e-02 -2.49107629e-01 6.28405571e-01
-1.03821611e+00 -5.22059917e-01 2.17779547e-01 3.31907958e-01
1.18752563e+00 1.91161573e-01 -3.07761222e-01 4.49898541e-01
7.19641089e-01 6.49648786e-01 3.27031016e-01 -8.49827468e-01
-4.99804989e-02 9.09303010e-01 -3.27606648e-02 -8.71647716e-01
-1.55929446e-01 -7.05526292e-01 -1.18651056e+00 9.09861863e-01
-1.33984506e-01 -9.65120271e-03 -8.75517011e-01 1.49117029e+00
2.18526617e-01 -3.03454399e-01 5.17811477e-01 5.97974002e-01
1.17246473e+00 6.07071400e-01 9.22391713e-02 -5.95637619e-01
1.00882816e+00 -7.51985013e-01 -9.22408700e-01 -9.15351231e-03
3.92648309e-01 -5.64058363e-01 8.56770158e-01 2.24378631e-01
-9.98007476e-01 -9.50469613e-01 -1.11181593e+00 7.67329857e-02
-2.78643876e-01 3.28627706e-01 5.93684912e-01 8.26558352e-01
-1.14564478e+00 4.72921044e-01 -5.01194298e-01 -2.19952613e-01
3.92211109e-01 4.11553502e-01 -6.56509876e-01 -3.77474539e-02
-9.99862313e-01 7.01497734e-01 4.56101507e-01 2.47919291e-01
-6.09680951e-01 -6.85698569e-01 -8.77379477e-01 7.69434720e-02
8.68022442e-02 -9.53129292e-01 6.12176120e-01 -9.83181238e-01
-1.43709946e+00 9.10908222e-01 1.24749970e-02 -2.03747407e-01
3.10643464e-01 -2.03009084e-01 -4.75883752e-01 4.88445133e-01
-1.88435882e-01 3.22938949e-01 1.25852895e+00 -1.81427681e+00
-4.28092957e-01 -4.08251047e-01 -7.92866647e-02 2.33654365e-01
-9.30348933e-01 -2.08210245e-01 1.23954333e-01 -4.77542073e-01
3.48679811e-01 -3.37395579e-01 -1.36884212e-01 4.51929063e-01
-1.75134823e-01 -1.78883225e-01 1.20119381e+00 -6.37902200e-01
1.02951574e+00 -2.44609976e+00 -8.49067494e-02 3.22619788e-02
4.87088740e-01 3.53721559e-01 7.99586922e-02 5.10649443e-01
-1.39864743e-01 -1.02955557e-01 -3.07420254e-01 -3.92071277e-01
-1.60427928e-01 -1.59461260e-01 -2.02324271e-01 7.12322712e-01
4.61883023e-02 7.71644890e-01 -6.91678464e-01 -6.03245676e-01
6.87211990e-01 8.12859774e-01 -1.09038316e-01 4.07801092e-01
1.62162095e-01 6.34747624e-01 -4.19023752e-01 7.48992682e-01
7.08778858e-01 -1.38654783e-01 -9.65919718e-02 -7.22039640e-01
-1.98186159e-01 -3.71220618e-01 -8.85281980e-01 1.75044215e+00
-4.61378843e-01 2.62085944e-01 1.01947568e-01 -1.18691337e+00
1.12437868e+00 4.24979746e-01 9.32724535e-01 -9.14485693e-01
2.85357982e-01 6.67737797e-03 -6.32081568e-01 -3.85574162e-01
1.27806351e-01 -3.32863182e-01 2.31744841e-01 3.12669784e-01
2.57233888e-01 1.01185486e-01 -1.73930824e-01 1.25263095e-01
8.32742691e-01 7.45802447e-02 8.14772174e-02 1.61822792e-02
7.95745969e-01 -2.86853462e-01 4.02880281e-01 5.49114704e-01
-1.39528159e-02 3.83857101e-01 -2.63133466e-01 -3.62811327e-01
-7.51333475e-01 -1.15228760e+00 -2.19701119e-02 8.91362250e-01
3.38629067e-01 -1.52421921e-01 -3.96043211e-01 -6.38899326e-01
-9.98849720e-02 3.70642573e-01 -4.73155379e-01 -5.79775631e-01
1.14046605e-02 -5.84787428e-01 1.05321907e-01 3.13793600e-01
7.00251222e-01 -1.04540122e+00 -6.43770635e-01 1.18995093e-01
-5.41045927e-02 -9.12074089e-01 7.28641525e-02 2.72827297e-01
-1.07659507e+00 -1.05701959e+00 -6.13145411e-01 -5.84750473e-01
7.05417335e-01 8.36283267e-01 6.84806168e-01 3.81303757e-01
-4.02151763e-01 8.13674092e-01 -5.56809664e-01 -1.06125280e-01
-4.86426413e-01 -2.46243551e-01 -5.45907766e-02 2.75080502e-01
1.73394769e-01 -5.84121048e-01 -8.13975990e-01 -1.64801050e-02
-1.17788208e+00 1.18006237e-01 7.58484542e-01 8.43500555e-01
3.18955451e-01 4.17059243e-01 2.92764008e-01 -2.68622458e-01
2.74320662e-01 -2.58005798e-01 -3.36387366e-01 4.30162758e-01
-9.01263356e-01 2.19052687e-01 4.63351160e-01 -7.91977253e-03
-1.36335444e+00 1.76844716e-01 -1.05401173e-01 -3.83267701e-01
-7.17774332e-02 2.29783610e-01 -1.46701813e-01 -4.41011846e-01
7.04399228e-01 5.66833675e-01 2.39801884e-01 -5.23546040e-01
4.75324750e-01 7.33767688e-01 5.37007570e-01 -1.91470206e-01
1.18320358e+00 9.04207587e-01 -4.52878736e-02 -7.90538192e-01
-8.06421340e-01 -6.38723195e-01 -4.97992039e-01 -4.60638613e-01
8.70356023e-01 -1.17117488e+00 -7.04423606e-01 5.75224936e-01
-7.84211397e-01 5.14703244e-02 -5.04674971e-01 5.64016998e-01
-5.97153783e-01 6.63634896e-01 -4.12476480e-01 -1.26715469e+00
-7.78191626e-01 -6.20020688e-01 9.90293622e-01 4.44633394e-01
3.81755888e-01 -1.03660309e+00 1.43697768e-01 5.54528892e-01
4.63389456e-01 4.78244931e-01 5.60216367e-01 -7.57086128e-02
-3.91310066e-01 -2.17551291e-01 -2.96761274e-01 7.09959626e-01
4.49465901e-01 2.62350645e-02 -1.30978966e+00 -6.04556620e-01
4.27334249e-01 -4.08292711e-01 1.41220832e+00 1.86456069e-01
8.33318174e-01 -2.32147887e-01 -2.38853276e-01 7.12523580e-01
1.60583079e+00 5.53524643e-02 8.05766940e-01 3.72727364e-01
5.10300994e-01 6.41952515e-01 5.63315868e-01 4.67858046e-01
1.50577292e-01 1.15987249e-01 6.37758493e-01 -8.19883466e-01
-1.84438229e-01 -4.30800766e-02 3.67633998e-01 6.38345897e-01
-2.63302296e-01 5.51542453e-02 -4.69442606e-01 2.87881941e-01
-1.84455633e+00 -1.14290202e+00 -1.75576769e-02 2.20315599e+00
8.78784537e-01 -7.06090257e-02 -1.49482206e-01 7.20925033e-01
4.30948138e-01 3.72764766e-01 -2.90444404e-01 2.52870053e-01
-2.77943015e-01 1.04022756e-01 3.39614332e-01 2.96042800e-01
-1.14554727e+00 5.47536671e-01 6.40453577e+00 7.13785172e-01
-1.13715053e+00 1.98901653e-01 4.08417434e-01 3.14639032e-01
-6.32246375e-01 5.24178669e-02 -4.81204271e-01 3.88815582e-01
6.92494571e-01 4.54008915e-02 -4.41927239e-02 3.84522259e-01
2.38170937e-01 -2.45999619e-01 -6.66740119e-01 1.19790411e+00
-2.09669843e-02 -1.17623305e+00 -1.62222445e-01 4.47227396e-02
6.92306161e-01 -3.09950262e-01 3.16517651e-01 -1.27127394e-01
9.26784649e-02 -8.34685743e-01 4.36372548e-01 9.91561532e-01
7.13502944e-01 -6.82649612e-01 7.52524972e-01 2.25388318e-01
-1.17120540e+00 -4.58052695e-01 -2.98444778e-01 2.43763939e-01
-8.36126581e-02 1.01190794e+00 5.27520813e-02 1.13469958e+00
7.69522429e-01 7.95113981e-01 -5.60602069e-01 9.34509397e-01
-1.32216170e-01 3.73159826e-01 -6.01799078e-02 5.79327643e-01
4.33454402e-02 -1.80156499e-01 3.10603201e-01 8.69868398e-01
2.64036536e-01 3.28630470e-02 1.09752879e-01 8.06998372e-01
1.67330712e-01 9.08678398e-02 -9.10237730e-01 -1.09160006e-01
-1.42946485e-02 1.39443767e+00 -4.03058589e-01 -2.32803047e-01
-4.26886618e-01 1.09690344e+00 -8.34060013e-02 4.64487582e-01
-6.44803405e-01 -2.76234269e-01 3.85859042e-01 6.93974569e-02
2.33257741e-01 -3.11120540e-01 -3.52760673e-01 -1.15738726e+00
-9.12596136e-02 -4.26688373e-01 3.28830481e-01 -8.12684357e-01
-1.26313400e+00 4.42408293e-01 -2.42418855e-01 -1.45468199e+00
2.18506306e-01 -3.19100648e-01 -6.93677187e-01 6.07664406e-01
-2.03982353e+00 -1.38156343e+00 -7.43918359e-01 9.70595419e-01
4.22264636e-02 -3.58190954e-01 7.34355152e-01 1.32199928e-01
-2.66941160e-01 4.96468276e-01 5.16313553e-01 -2.52957255e-01
6.94192946e-01 -8.46836388e-01 -4.83142525e-01 8.16476882e-01
2.36787498e-01 2.65714586e-01 6.51355147e-01 -3.46278071e-01
-1.55096114e+00 -8.81388247e-01 5.21189690e-01 4.05925289e-02
2.57051587e-01 -2.42381152e-02 -8.96900475e-01 1.22826666e-01
5.09616971e-01 2.17740700e-01 9.23529029e-01 -2.21508473e-01
-5.31632304e-01 -6.96470022e-01 -1.34626794e+00 1.34430960e-01
8.15087080e-01 -7.66349256e-01 -5.05379498e-01 3.73902142e-01
7.40118444e-01 2.91612834e-01 -1.02609313e+00 6.42446518e-01
3.99937749e-01 -1.31307137e+00 1.19242907e+00 -1.36364967e-01
2.42752880e-01 -4.11542475e-01 -1.35041595e-01 -9.78642523e-01
-4.30848330e-01 -5.20146072e-01 -3.13244909e-01 1.29580343e+00
-7.46782199e-02 -6.26479506e-01 5.09455621e-01 2.01965645e-01
9.74448919e-02 -3.49682897e-01 -7.88033545e-01 -7.31533527e-01
-3.34672660e-01 4.84744795e-02 -6.79385662e-03 9.27819908e-01
-1.55230105e-01 1.39295369e-01 -4.17520940e-01 1.19567558e-01
1.19532967e+00 6.47775270e-03 5.76712728e-01 -1.48389745e+00
9.41353675e-04 -2.70897537e-01 -4.83501285e-01 -2.82973468e-01
1.44737154e-01 -6.46578431e-01 -9.85293686e-02 -1.81896579e+00
3.67134601e-01 -2.56141841e-01 -9.06595230e-01 7.46600270e-01
-1.92957997e-01 5.42135119e-01 2.06177950e-01 3.20328176e-01
-5.42741835e-01 9.99944508e-01 1.06112635e+00 -4.79412436e-01
-1.86587855e-01 -1.60579771e-01 -9.63916719e-01 4.39232409e-01
1.05510628e+00 -3.41847062e-01 -4.38820869e-01 -2.34173447e-01
2.52316315e-02 -1.28814742e-01 6.42731607e-01 -1.21079743e+00
4.57161903e-01 -1.38968408e-01 5.04563510e-01 -5.12683034e-01
3.40411007e-01 -1.12836289e+00 1.58580393e-01 5.98838747e-01
-1.34414872e-02 -6.67699635e-01 3.89610380e-02 7.08446741e-01
-3.53250831e-01 -4.74524638e-03 8.72627378e-01 -4.89593949e-03
-5.71330011e-01 1.50685906e-01 -1.55471653e-01 -6.45608962e-01
1.05005777e+00 -5.88424623e-01 -2.42478788e-01 -2.60227591e-01
-7.49112666e-01 -7.96456411e-02 4.99053419e-01 2.91946203e-01
8.23051095e-01 -1.39730370e+00 -6.35497510e-01 3.01295698e-01
2.84481466e-01 -3.41662765e-01 4.50680733e-01 9.44087327e-01
-1.53980032e-02 3.12058302e-03 -3.17760974e-01 -5.21773458e-01
-1.42438149e+00 7.15338767e-01 1.43878743e-01 -2.52094626e-01
-7.69530177e-01 4.51971352e-01 9.21971872e-02 -2.57046372e-01
2.27706999e-01 1.08848572e-01 -2.63442367e-01 2.68754736e-02
5.88341296e-01 3.28636676e-01 -4.74696495e-02 -5.56770444e-01
-2.29696974e-01 8.63666654e-01 2.65886158e-01 1.55323550e-01
1.46205592e+00 -3.88683736e-01 -1.98553532e-01 2.08305702e-01
1.22729027e+00 -1.03626430e-01 -1.22175419e+00 -4.51427609e-01
-4.38492239e-01 -4.39609468e-01 5.50659835e-01 -6.30195737e-01
-1.26369369e+00 8.89478087e-01 1.09612656e+00 2.16014788e-01
1.64415467e+00 -8.17688182e-02 6.30042255e-01 3.32888901e-01
2.17577010e-01 -9.95553613e-01 4.66177553e-01 -1.56605557e-01
6.72751307e-01 -1.47548151e+00 1.61608487e-01 -1.90350547e-01
-2.90984988e-01 1.22593820e+00 2.09030554e-01 1.63465962e-01
6.24036252e-01 1.71784073e-01 1.44962475e-01 -2.19285131e-01
-3.79956722e-01 -5.58179677e-01 3.89110297e-01 7.61815310e-01
3.91410470e-01 -3.95086348e-01 -2.28042424e-01 1.74541384e-01
4.39178526e-01 1.06287263e-01 -7.42212385e-02 1.16804767e+00
-9.83950794e-01 -9.92118657e-01 -6.64341688e-01 7.89626613e-02
-2.43151769e-01 9.47610736e-02 -1.88303381e-01 3.32033813e-01
1.74667746e-01 1.18266821e+00 -2.50631124e-01 -5.28297305e-01
-7.03237429e-02 1.38754575e-02 5.10006070e-01 -1.79092661e-01
-5.17743647e-01 1.54981703e-01 -4.00960952e-01 -3.45607400e-01
-1.11582267e+00 -2.65537072e-02 -1.02045250e+00 -2.46352404e-01
-5.14646828e-01 6.45054951e-02 7.66205490e-01 8.38842154e-01
2.47313246e-01 5.26697338e-01 1.15171659e+00 -7.29226351e-01
-3.53722870e-01 -7.15047598e-01 -6.25322878e-01 3.42399180e-01
6.02848828e-01 -5.36776662e-01 -7.11567342e-01 1.56132892e-01] | [10.488018989562988, -1.9987648725509644] |
dd387519-ba00-42e0-901d-1bb2ffba2eac | generating-high-quality-emotion-arcs-for-low | 2306.02213 | null | https://arxiv.org/abs/2306.02213v1 | https://arxiv.org/pdf/2306.02213v1.pdf | Generating High-Quality Emotion Arcs For Low-Resource Languages Using Emotion Lexicons | Automatically generated emotion arcs -- that capture how an individual or a population feels over time -- are widely used in industry and research. However, there is little work on evaluating the generated arcs in English (where the emotion resources are available) and no work on generating or evaluating emotion arcs for low-resource languages. Work on generating emotion arcs in low-resource languages such as those indigenous to Africa, the Americas, and Australia is stymied by the lack of emotion-labeled resources and large language models for those languages. Work on evaluating emotion arcs (for any language) is scarce because of the difficulty of establishing the true (gold) emotion arc. Our work, for the first time, systematically and quantitatively evaluates automatically generated emotion arcs. We also compare two common ways of generating emotion arcs: Machine-Learning (ML) models and Lexicon-Only (LexO) methods. By running experiments on 42 diverse datasets in 9 languages, we show that despite being markedly poor at instance level emotion classification, LexO methods are highly accurate at generating emotion arcs when aggregating information from hundreds of instances. (Predicted arcs have correlations ranging from 0.94 to 0.99 with the gold arcs for various emotions.) We also show that for languages with no emotion lexicons, automatic translations of English emotion lexicons can be used to generate high-quality emotion arcs -- correlations above 0.9 with the gold emotion arcs in all six indigenous African languages explored. This opens up avenues for work on emotions in numerous languages from around the world; crucial not only for commerce, public policy, and health research in service of speakers of those languages, but also to draw meaningful conclusions in emotion-pertinent research using information from around the world (thereby avoiding a western-centric bias in research). | ['Saif M. Mohammad', 'Daniela Teodorescu'] | 2023-06-03 | null | null | null | null | ['emotion-classification', 'emotion-classification'] | ['computer-vision', 'natural-language-processing'] | [ 3.42965536e-02 2.02053174e-01 -4.59721982e-01 -5.19478202e-01
-8.20244491e-01 -7.09082603e-01 4.77231175e-01 1.85406849e-01
-4.30815428e-01 8.56422722e-01 5.13790071e-01 -4.74486798e-01
1.96013168e-01 -6.48375094e-01 -1.07205406e-01 -2.98127353e-01
1.16644002e-01 4.05614913e-01 -6.36171877e-01 -5.08560240e-01
7.07315281e-02 5.56697667e-01 -1.27673769e+00 1.53832197e-01
9.78230178e-01 7.13029921e-01 -5.23803413e-01 4.72664863e-01
-4.86891180e-01 8.85558903e-01 -7.85666943e-01 -7.31340289e-01
-1.16122440e-01 -8.45394433e-01 -8.36625814e-01 -1.37868881e-01
-6.53865114e-02 2.84818947e-01 4.38131005e-01 8.74649525e-01
4.45072919e-01 -1.47759337e-02 6.07564926e-01 -1.44980681e+00
-8.35995317e-01 5.16786873e-01 -3.28469247e-01 -2.12413684e-01
6.32999778e-01 -4.13471237e-02 1.00888276e+00 -5.28699815e-01
9.67499554e-01 1.10643840e+00 7.21664429e-01 7.31310666e-01
-9.73278046e-01 -8.15485835e-01 7.07594603e-02 -2.70728856e-01
-1.03042769e+00 -2.47155920e-01 7.97920346e-01 -2.97650695e-01
1.43628418e+00 5.31371891e-01 8.11285555e-01 1.49708343e+00
2.92782396e-01 2.73434073e-01 1.66125584e+00 -4.17289019e-01
3.51761989e-02 5.91747880e-01 -1.73411623e-01 4.60442215e-01
-1.72803968e-01 -1.88629657e-01 -3.68852258e-01 -2.32704371e-01
5.64350009e-01 -4.66203034e-01 1.95690259e-01 5.25572717e-01
-1.06173062e+00 7.77895153e-01 8.61558616e-02 6.22116685e-01
-4.98863846e-01 -8.30339864e-02 7.11455524e-01 5.56221008e-01
1.08382976e+00 7.77251184e-01 -6.59083128e-01 -8.40306044e-01
-6.69772387e-01 8.38342775e-03 1.21984100e+00 6.79490089e-01
6.68980420e-01 3.69560540e-01 5.56013942e-01 1.31770122e+00
-1.36264833e-02 3.67764711e-01 4.50295717e-01 -1.06784511e+00
1.47286564e-01 6.22968793e-01 1.39817864e-01 -1.14602125e+00
-6.88563228e-01 -1.24666132e-01 -6.17222071e-01 1.91427663e-01
2.92758197e-01 -8.23514462e-01 -5.08463085e-01 1.97313547e+00
1.37770414e-01 -2.47191310e-01 4.75405037e-01 7.41471708e-01
7.29713082e-01 8.44470263e-01 5.66641092e-01 -4.45217788e-01
1.52517951e+00 -6.77445889e-01 -7.33907640e-01 -4.67854530e-01
8.91677260e-01 -8.91845703e-01 1.38315892e+00 5.36360323e-01
-1.00848877e+00 -7.54413754e-02 -7.01249003e-01 1.40403032e-01
-8.37346554e-01 -9.27512348e-02 1.17098486e+00 8.94710302e-01
-1.12450349e+00 1.26709742e-02 -4.90869850e-01 -6.61805212e-01
2.34512702e-01 1.66457683e-01 -5.35804093e-01 2.91403025e-01
-1.35751188e+00 1.38255906e+00 -1.26380518e-01 4.78437319e-02
-9.21217799e-02 -5.57942450e-01 -8.53432953e-01 -2.92427778e-01
4.00308520e-02 -9.76073220e-02 9.79722321e-01 -2.05810237e+00
-1.45558631e+00 1.08291626e+00 -1.97668090e-01 1.16968460e-01
6.38834611e-02 4.57446510e-03 -1.09343171e+00 -8.81049410e-02
1.84965447e-01 6.69981420e-01 1.42640933e-01 -1.20149827e+00
-4.93347883e-01 -1.81769967e-01 -8.59134719e-02 3.02779108e-01
-4.44211096e-01 8.60024393e-01 -1.02137342e-01 -6.92710876e-01
-3.43248874e-01 -9.21537042e-01 -4.57614034e-01 -5.46810567e-01
-1.14806776e-03 -4.54841137e-01 3.47462744e-01 -8.17968607e-01
1.10387349e+00 -1.96964061e+00 -2.86783457e-01 4.58750784e-01
-2.60742784e-01 -2.14946449e-01 -2.67318785e-01 6.17048264e-01
-2.99795151e-01 5.99354088e-01 -1.11905195e-01 1.55160129e-01
1.76805079e-01 4.18950707e-01 -3.17435116e-01 2.28945822e-01
6.19074762e-01 8.59729886e-01 -1.00079846e+00 -4.75079149e-01
-2.56120674e-02 6.23908222e-01 -4.38893765e-01 -1.66898564e-01
-2.06876136e-02 8.12581554e-02 -2.37644956e-01 7.94514835e-01
9.65625234e-03 2.25421205e-01 3.02498698e-01 2.14233637e-01
-3.29751819e-01 2.41447866e-01 -8.15981865e-01 1.34913790e+00
-9.16264832e-01 8.46596122e-01 -7.80061930e-02 -7.47813046e-01
1.24261796e+00 5.65989435e-01 8.09283674e-01 -9.40029085e-01
1.37522265e-01 4.15063500e-01 1.08396411e-01 -6.14128709e-01
5.71216464e-01 -6.83120787e-01 -6.52801454e-01 8.12330544e-01
-7.98692331e-02 -4.48597193e-01 1.03156842e-01 1.91394333e-02
8.93254578e-01 8.72533694e-02 1.46714628e-01 -3.21525484e-01
4.72492613e-02 3.68113071e-01 5.64813375e-01 1.76035240e-01
-1.87230438e-01 6.32581860e-02 8.57831419e-01 -3.64826262e-01
-1.01903534e+00 -7.28853822e-01 -3.84881377e-01 1.28197014e+00
-3.86171967e-01 -4.30780917e-01 -5.93925297e-01 -3.56369108e-01
-5.70690334e-01 8.21995497e-01 -5.97996056e-01 1.31784696e-02
-2.84624755e-01 -9.65291798e-01 9.88840401e-01 5.37108302e-01
-5.83796948e-02 -1.63971591e+00 -7.28756905e-01 3.50221545e-01
-4.11845475e-01 -1.16031587e+00 6.18827976e-02 2.62781084e-01
-4.80058193e-01 -6.58120871e-01 -5.19455612e-01 -8.78134191e-01
5.77236116e-01 -7.25359440e-01 1.33063245e+00 -2.54147112e-01
-3.46796870e-01 5.73086917e-01 -4.34262812e-01 -8.31024647e-01
-5.59268653e-01 -2.03013763e-01 -4.48390432e-02 -3.14384967e-01
8.17150235e-01 -3.40047121e-01 1.93228684e-02 1.02961548e-01
-7.94945896e-01 -2.24755973e-01 4.46971327e-01 5.47702372e-01
2.56590962e-01 8.18998665e-02 1.18163764e+00 -1.02821827e+00
1.17072773e+00 -8.70500922e-01 -7.14261259e-05 1.67421997e-01
-6.78385437e-01 -3.17407459e-01 5.15076756e-01 -6.39930665e-01
-1.13831878e+00 -2.96377808e-01 -4.42412317e-01 3.13273072e-01
-4.71226066e-01 6.00868821e-01 1.76715925e-01 2.69170910e-01
7.76149333e-01 -4.84854639e-01 -2.54589170e-01 4.27558459e-02
4.32210594e-01 8.89429927e-01 4.88798916e-01 -9.64379251e-01
7.06814304e-02 1.58385813e-01 -3.55644464e-01 -9.44458663e-01
-4.02652383e-01 -2.02929184e-01 -6.21295683e-02 -4.95436907e-01
9.54651117e-01 -9.00897980e-01 -6.00624681e-01 1.42161712e-01
-9.37536657e-01 -6.78695500e-01 -3.61162007e-01 5.83269596e-01
-5.32412946e-01 -2.25451857e-01 -8.36285055e-01 -9.80903387e-01
-2.75941849e-01 -9.38151181e-01 9.24877167e-01 2.66040802e-01
-1.08793664e+00 -1.36022949e+00 3.26680273e-01 2.60758828e-02
3.68321359e-01 6.66354954e-01 1.00034010e+00 -7.47891665e-01
6.15255296e-01 -1.66304260e-01 -7.89828300e-02 3.65044057e-01
2.83929765e-01 3.66769940e-01 -8.11900973e-01 3.91104430e-01
6.52830070e-03 -8.43411684e-01 7.30146468e-02 6.98809624e-02
6.03845656e-01 -4.09390688e-01 1.16280690e-01 4.00900766e-02
1.38599396e+00 4.05610383e-01 6.41590595e-01 3.29453588e-01
3.72139007e-01 1.09490716e+00 6.64774954e-01 3.46909106e-01
6.56790674e-01 3.15685451e-01 -1.15740046e-01 -4.38689947e-01
4.43844020e-01 1.53434232e-01 7.37301290e-01 9.80354071e-01
-2.57097751e-01 -2.29933724e-01 -1.06793368e+00 7.92841613e-01
-1.53584480e+00 -1.03659582e+00 -2.31900692e-01 1.69303703e+00
1.18148100e+00 -1.30853262e-02 1.63491353e-01 -1.18215188e-01
2.36713961e-01 9.27002653e-02 -2.94573754e-01 -1.71332669e+00
-3.38357717e-01 7.57246315e-01 -2.21482106e-02 6.96480095e-01
-6.45778537e-01 1.13831830e+00 6.79859257e+00 4.49790061e-01
-1.34275973e+00 -4.67117094e-02 8.52641821e-01 -2.10409135e-01
-6.67702675e-01 9.67062861e-02 -1.70597136e-01 8.74757841e-02
1.28009999e+00 -3.00055623e-01 5.73613226e-01 7.22299814e-01
2.33269870e-01 -4.96439077e-02 -7.67707527e-01 8.50960076e-01
2.32413948e-01 -6.86069250e-01 -4.46300119e-01 -1.97743431e-01
9.28330898e-01 3.68917361e-02 -7.15599656e-02 4.72230017e-01
6.34831965e-01 -1.29622769e+00 6.83738589e-01 4.01083350e-01
9.18846846e-01 -1.15067565e+00 6.84694111e-01 -7.90659222e-04
-8.67932320e-01 2.43784621e-01 -2.01586097e-01 -3.38914305e-01
1.75260037e-01 6.53950334e-01 -5.11394203e-01 2.41177320e-01
8.29565704e-01 6.50686204e-01 -3.04755688e-01 4.27822843e-02
-2.62430936e-01 7.93815851e-01 -3.36384952e-01 -2.83568084e-01
6.08873785e-01 -4.98944163e-01 1.44309148e-01 1.83868682e+00
2.49608710e-01 2.61545092e-01 -1.19942583e-01 6.18271589e-01
-9.70629677e-02 7.04526126e-01 -7.66223550e-01 -5.00201046e-01
2.82579213e-01 1.66230011e+00 -9.07534540e-01 -2.11481035e-01
-5.85596681e-01 9.52760994e-01 2.72040039e-01 5.41418076e-01
-8.14113200e-01 -4.92000252e-01 7.70082057e-01 -3.41767818e-01
-5.14075756e-01 7.02904388e-02 -3.24217200e-01 -8.75050604e-01
-8.96640196e-02 -1.39069092e+00 3.44852775e-01 -1.02734530e+00
-1.58912313e+00 8.20216417e-01 -1.19095773e-01 -5.90197206e-01
-6.07308507e-01 -6.77932262e-01 -5.76030135e-01 9.51659799e-01
-9.07450497e-01 -1.07024705e+00 6.96228072e-02 4.41973954e-01
2.85596609e-01 1.42169029e-01 1.20632279e+00 1.75626993e-01
-4.00027961e-01 4.33866173e-01 -3.39461952e-01 1.35479346e-01
9.15834188e-01 -1.26280308e+00 1.32743150e-01 3.85172904e-01
2.66224653e-01 5.67584813e-01 5.60237885e-01 -6.58123970e-01
-1.00329626e+00 -6.34379923e-01 1.40620124e+00 -5.81122458e-01
1.01982653e+00 -3.16540450e-01 -7.54076362e-01 8.46511900e-01
6.71127796e-01 -5.02212346e-01 1.26743782e+00 6.02212906e-01
-3.38574439e-01 1.35349259e-01 -1.09900582e+00 1.03423822e+00
6.46753967e-01 -7.56608129e-01 -3.91118467e-01 2.51729846e-01
4.35256869e-01 -6.02110960e-02 -1.41993773e+00 1.56458512e-01
8.74277711e-01 -7.88521886e-01 5.70536852e-01 -9.81890082e-01
7.83313751e-01 2.91351199e-01 -9.84067768e-02 -1.63366771e+00
1.44146923e-02 -5.75290024e-01 7.35268176e-01 1.64035857e+00
8.16146433e-01 -9.93610263e-01 4.25376505e-01 1.01338446e+00
-2.00199693e-01 -1.07923293e+00 -6.03780329e-01 -3.86367559e-01
4.42482024e-01 -9.69947696e-01 6.04271829e-01 1.54443288e+00
5.74863315e-01 3.98162037e-01 -1.30820096e-01 -2.96878874e-01
-1.20680764e-01 -2.86466740e-02 5.86548328e-01 -9.86544967e-01
5.17973825e-02 -8.30721438e-01 -2.24810779e-01 2.50598602e-02
6.70500040e-01 -8.40710521e-01 1.16329506e-01 -1.62862372e+00
3.55426775e-04 -7.29284465e-01 -6.06733747e-02 8.85698020e-01
-1.45346299e-02 4.89268780e-01 -1.17007092e-01 -2.61249989e-01
-1.18368052e-01 3.11230980e-02 1.00176585e+00 3.77974540e-01
-2.87788242e-01 -6.15718722e-01 -1.28442872e+00 9.29953754e-01
1.12724924e+00 -5.51877320e-01 -5.10348558e-01 -1.84582204e-01
8.92462552e-01 -8.13741162e-02 1.95938870e-01 -4.93594557e-01
-2.34172717e-01 -5.53987801e-01 2.95066655e-01 3.67066152e-02
1.53018475e-01 -7.77551115e-01 4.26626205e-01 1.65646598e-02
-4.09967989e-01 6.15717947e-01 5.54329276e-01 -1.69914410e-01
-1.72657326e-01 -1.79840639e-01 4.79762197e-01 -1.68730497e-01
-5.67037404e-01 -1.32819459e-01 -7.42776752e-01 5.03199577e-01
9.02805209e-01 -3.40477638e-02 -3.51683885e-01 -6.47916734e-01
-5.85840821e-01 -1.63581762e-02 6.08183205e-01 6.55339420e-01
2.73520917e-01 -1.38346505e+00 -8.23300540e-01 1.87164433e-02
7.73798972e-02 -5.50048053e-01 -1.33809313e-01 7.78718531e-01
-4.25334781e-01 1.12539850e-01 -4.80855793e-01 2.00839117e-02
-8.88296068e-01 2.46349737e-01 2.19099432e-01 -2.90308714e-01
-1.73594072e-01 4.97341424e-01 -2.15681806e-01 -6.38217568e-01
-8.31873864e-02 -2.50795513e-01 4.10907157e-03 3.58061552e-01
1.63156882e-01 2.12787628e-01 -2.07198843e-01 -1.00018382e+00
-4.39416885e-01 4.06137764e-01 4.33466196e-01 -6.82148516e-01
1.40465438e+00 1.85364321e-01 -4.80448574e-01 9.34121847e-01
1.11611772e+00 4.88005489e-01 -5.20271659e-01 5.94251335e-01
9.01842769e-03 1.33217350e-01 -2.62543201e-01 -1.02309179e+00
-7.91478455e-01 7.27430999e-01 1.07207999e-01 3.32002342e-01
1.33713663e+00 4.37749512e-02 5.91249943e-01 1.16497524e-01
2.10729152e-01 -1.55543673e+00 -1.84777565e-02 3.23884517e-01
8.91243696e-01 -8.87022913e-01 -2.43571728e-01 -4.47605997e-02
-1.35876250e+00 8.90224040e-01 5.75222790e-01 8.48968886e-03
4.57210600e-01 6.34235978e-01 7.89630890e-01 -3.92265618e-01
-8.54702294e-01 -6.86524361e-02 7.01458612e-03 5.96033156e-01
1.01942086e+00 3.93347472e-01 -6.18109524e-01 8.89656842e-01
-5.02815068e-01 -1.41068131e-01 4.59558129e-01 6.64614618e-01
2.73657497e-02 -1.35909534e+00 -2.72483706e-01 6.61948025e-01
-8.49629283e-01 -2.41919339e-01 -1.09354937e+00 1.07103002e+00
2.44636193e-01 1.17786896e+00 2.13152558e-01 -2.50420868e-01
2.15510875e-01 4.86039132e-01 2.74946272e-01 -3.72683287e-01
-1.05904210e+00 1.51250541e-01 7.83468246e-01 -4.84309077e-01
-7.27435887e-01 -6.77728474e-01 -1.55442023e+00 -3.05179656e-01
-1.02922827e-01 3.95934880e-01 7.81511664e-01 8.16862166e-01
8.69427323e-02 3.27616870e-01 5.54658473e-01 -5.23054600e-01
2.45339394e-01 -7.84587741e-01 -5.82861066e-01 5.39485812e-01
-3.89351100e-01 -1.71437129e-01 -3.29627723e-01 2.40344368e-03] | [12.823442459106445, 6.244893550872803] |
9b8a8b10-a751-40c7-850b-cb873e0615da | multi-head-uncertainty-inference-for | 2212.10006 | null | https://arxiv.org/abs/2212.10006v1 | https://arxiv.org/pdf/2212.10006v1.pdf | Multi-head Uncertainty Inference for Adversarial Attack Detection | Deep neural networks (DNNs) are sensitive and susceptible to tiny perturbation by adversarial attacks which causes erroneous predictions. Various methods, including adversarial defense and uncertainty inference (UI), have been developed in recent years to overcome the adversarial attacks. In this paper, we propose a multi-head uncertainty inference (MH-UI) framework for detecting adversarial attack examples. We adopt a multi-head architecture with multiple prediction heads (i.e., classifiers) to obtain predictions from different depths in the DNNs and introduce shallow information for the UI. Using independent heads at different depths, the normalized predictions are assumed to follow the same Dirichlet distribution, and we estimate distribution parameter of it by moment matching. Cognitive uncertainty brought by the adversarial attacks will be reflected and amplified on the distribution. Experimental results show that the proposed MH-UI framework can outperform all the referred UI methods in the adversarial attack detection task with different settings. | ['Kongming Liang', 'Ke Zhang', 'Kai Guo', 'Jiyang Xie. Zhongwei Si', 'Songyun Yang', 'YuQi Yang'] | 2022-12-20 | null | null | null | null | ['adversarial-defense', 'adversarial-attack-detection', 'adversarial-attack-detection'] | ['adversarial', 'computer-vision', 'knowledge-base'] | [-2.98398640e-02 4.68290716e-01 3.99099201e-01 -5.36524653e-01
-7.60840356e-01 -6.61799431e-01 6.80144608e-01 -1.47929639e-01
-1.82351544e-01 9.73200738e-01 4.78082478e-01 2.58933585e-02
7.26004094e-02 -1.04855251e+00 -9.07800555e-01 -8.79366279e-01
1.19579278e-01 5.46164930e-01 4.25580144e-01 -5.95144331e-02
1.11861534e-01 3.99988651e-01 -1.15671599e+00 2.50500888e-01
7.19419599e-01 1.15484381e+00 -2.91246235e-01 4.88821268e-01
-6.04841597e-02 9.16257977e-01 -1.34900045e+00 -1.09231472e+00
1.21138781e-01 6.35892004e-02 -3.36020052e-01 -6.01095319e-01
1.09846527e-02 -7.68579483e-01 -6.18530810e-01 1.69043028e+00
9.14313912e-01 1.63178086e-01 1.02663076e+00 -1.52838457e+00
-7.76502013e-01 1.35135770e+00 -5.11232793e-01 1.93653330e-01
7.98361003e-02 -2.18911573e-01 3.12680125e-01 -6.88048005e-01
5.02879396e-02 2.07621884e+00 7.36124992e-01 1.05859208e+00
-5.05561292e-01 -1.22679055e+00 3.82304043e-01 4.36947614e-01
-1.37266552e+00 -5.25101006e-01 1.03504705e+00 -1.99320212e-01
3.58920157e-01 -1.54229195e-03 -2.64790773e-01 2.01226521e+00
4.59339917e-01 9.07362998e-01 7.95384586e-01 -8.93194824e-02
7.74716020e-01 1.69306606e-01 -1.58551738e-01 1.59443513e-01
2.24386394e-01 1.88875735e-01 -1.99677199e-01 -5.89305401e-01
3.52281034e-01 2.07996666e-01 -2.29827881e-01 3.90827537e-01
-3.05651277e-01 9.96829748e-01 4.60426539e-01 -1.94914177e-01
-2.18402296e-01 1.04566231e-01 7.03787506e-01 7.72030577e-02
5.85656285e-01 -1.49091616e-01 -5.56840479e-01 1.17318511e-01
-4.14648682e-01 3.13650012e-01 1.01916492e+00 1.16179204e+00
1.84642628e-01 2.79411018e-01 -4.15033609e-01 5.19561827e-01
6.09952509e-01 6.11984074e-01 4.84194100e-01 -6.61086202e-01
4.84378874e-01 -5.92043027e-02 -5.31229191e-03 -1.02713501e+00
-2.01129064e-01 -4.14016157e-01 -1.25280333e+00 4.15766895e-01
2.45048448e-01 -9.40488756e-01 -1.01859403e+00 2.07608175e+00
3.26857418e-01 7.27895856e-01 4.58699137e-01 4.80398387e-01
8.05031121e-01 5.18016756e-01 3.99597079e-01 5.92393726e-02
9.60201025e-01 -3.73789102e-01 -7.07708776e-01 -3.20780426e-01
7.66038299e-02 -4.54821646e-01 4.22045112e-01 5.02668083e-01
-6.79671586e-01 -2.50107646e-01 -1.09404504e+00 4.79571521e-01
-3.83813649e-01 -6.47163868e-01 3.32728289e-02 1.07213366e+00
-5.16772151e-01 6.30493343e-01 -5.51936328e-01 4.45406556e-01
8.19125235e-01 1.57044590e-01 -4.26341631e-02 -1.81030199e-01
-2.08095193e+00 8.14485312e-01 7.23931372e-01 1.36369246e-03
-1.59632313e+00 -5.99059582e-01 -7.24092126e-01 5.62354550e-03
1.47997975e-01 -3.72858286e-01 1.09096849e+00 -7.77056515e-01
-1.37310243e+00 3.67354453e-01 2.35963449e-01 -6.58358216e-01
7.70171225e-01 -3.62783372e-01 -4.25638020e-01 -1.71298638e-01
-2.68267334e-01 2.62197018e-01 1.24152887e+00 -1.37868261e+00
-3.30688447e-01 -6.51715696e-01 7.43906498e-02 3.40114757e-02
-4.67941970e-01 3.65642250e-01 -4.30237167e-02 -8.88030827e-01
-1.39750004e-01 -6.97050393e-01 -2.59517759e-01 -2.62175471e-01
-7.88273096e-01 -1.41621783e-01 9.22987342e-01 -5.46914339e-01
1.07869625e+00 -2.07345748e+00 9.79535729e-02 2.13421822e-01
3.05686235e-01 4.67364371e-01 2.53332913e-01 5.49542569e-02
1.36720762e-01 8.94838795e-02 -3.79556805e-01 -5.30743301e-01
5.56645572e-01 6.44357502e-01 -9.95546877e-01 4.52141583e-01
-5.56269027e-02 4.66486931e-01 -7.58506894e-01 -3.31429303e-01
-5.91710396e-02 5.74575007e-01 -3.43292385e-01 4.00394410e-01
-3.13835949e-01 2.50466615e-01 -5.67470312e-01 7.69225240e-01
1.20267558e+00 6.50773227e-01 -3.60862315e-01 1.41613595e-02
7.07931519e-01 -3.44742537e-01 -1.25277114e+00 9.03122246e-01
-2.05923654e-02 2.62272060e-01 5.60924634e-02 -8.45263839e-01
1.10690689e+00 4.31952089e-01 -2.85292357e-01 2.40908433e-02
7.98394918e-01 -1.13659412e-01 -2.68719308e-02 -1.00093663e-01
-4.79161441e-02 -2.26664290e-01 -5.27199745e-01 1.57369018e-01
2.39401042e-01 2.53940523e-01 -6.83736980e-01 1.10723287e-01
1.19007707e+00 -4.17983800e-01 8.84806290e-02 -4.98413406e-02
6.44516051e-01 -7.94190288e-01 1.13419986e+00 1.13355875e+00
-7.55411327e-01 4.64561433e-01 7.24904060e-01 -2.58330405e-01
-8.92590165e-01 -1.56870019e+00 -4.60368365e-01 1.10681236e+00
9.43915024e-02 7.61929229e-02 -1.12750685e+00 -8.92693102e-01
8.86331573e-02 9.58572447e-01 -7.97152162e-01 -6.68701470e-01
-9.75218788e-02 -7.87324667e-01 1.19841075e+00 7.96200037e-01
9.58889902e-01 -1.02970123e+00 7.69377202e-02 3.27814817e-01
4.48803864e-02 -1.19756556e+00 9.71275419e-02 2.02987552e-01
-4.04117793e-01 -5.76568305e-01 -7.93193698e-01 -3.47070277e-01
4.04479831e-01 -5.47020376e-01 8.20376933e-01 -6.38487339e-01
3.69941182e-02 1.11833319e-01 -4.23856825e-01 -1.29305041e+00
-5.71434200e-01 -3.85833085e-01 7.77378619e-01 -1.27786085e-01
5.20054340e-01 -9.15580332e-01 -3.67033541e-01 2.36733794e-01
-1.17483962e+00 -9.08434689e-01 5.21543086e-01 5.06566942e-01
4.27801877e-01 5.12499690e-01 1.02853394e+00 -1.01192749e+00
9.11743462e-01 -1.11179447e+00 -2.15369165e-01 1.45486191e-01
-1.66252777e-01 1.09705560e-01 9.06074703e-01 -7.46459067e-01
-1.33167613e+00 -3.56475979e-01 -5.96857667e-01 -8.78644764e-01
-5.97937524e-01 2.01134935e-01 -1.12523544e+00 -9.25952345e-02
9.30260420e-01 -1.10944293e-01 -5.36601722e-01 -2.42132172e-01
3.89570534e-01 7.96695769e-01 8.85770500e-01 -6.96531415e-01
9.18263912e-01 1.80508837e-01 -2.96889357e-02 -3.48822087e-01
-1.12581158e+00 3.45704406e-01 -4.28434670e-01 -1.46311507e-01
6.88678086e-01 -9.09054041e-01 -4.45136130e-01 1.16628039e+00
-1.31327271e+00 9.45228264e-02 -1.07645784e-02 3.16709280e-01
-2.28625178e-01 4.27337736e-01 -7.34770894e-01 -1.20091307e+00
-6.34219587e-01 -1.00083673e+00 7.42894113e-01 5.12058735e-01
2.47911155e-01 -9.58294809e-01 -1.45636788e-02 -8.52028877e-02
3.89028549e-01 6.99753702e-01 7.96999991e-01 -1.44488990e+00
-7.75443614e-02 -5.63943744e-01 -1.56837910e-01 7.26345479e-01
-3.02532941e-01 -6.47897720e-02 -1.52696657e+00 5.77168949e-02
4.01494175e-01 -4.44406420e-01 8.29112887e-01 3.98733705e-01
1.44207525e+00 -5.85658550e-01 -2.83648342e-01 5.46885908e-01
1.04030752e+00 1.38320610e-01 8.65677416e-01 1.52924970e-01
7.15237916e-01 3.36460948e-01 3.19800824e-01 8.76337707e-01
1.51258186e-01 -1.57950297e-02 1.08880699e+00 6.04073286e-01
4.42994088e-01 -1.83396220e-01 6.74662054e-01 3.29354703e-01
2.77945429e-01 -6.29273295e-01 -9.44859445e-01 2.31461957e-01
-1.75658846e+00 -1.10341871e+00 3.07741195e-01 1.94325793e+00
9.68031466e-01 7.96637774e-01 -2.11344004e-01 2.25306600e-01
1.04989552e+00 2.10530803e-01 -1.02719092e+00 -5.19638360e-01
-1.13416195e-01 7.47543648e-02 3.46065372e-01 4.32792336e-01
-1.30513704e+00 7.88500607e-01 5.88370657e+00 1.04449344e+00
-4.04940963e-01 1.11041062e-01 8.68290961e-01 -4.58994247e-02
-3.58946741e-01 -5.70600331e-01 -1.08979619e+00 8.10363352e-01
1.07191372e+00 -3.41771007e-01 4.75799814e-02 1.25392342e+00
-4.95849371e-01 4.49077308e-01 -9.16911066e-01 7.07959294e-01
1.68329760e-01 -6.61429048e-01 2.53474146e-01 -1.87776640e-01
7.82899320e-01 -4.70249280e-02 4.73392695e-01 7.25547910e-01
1.00466812e+00 -9.79658246e-01 5.62838376e-01 9.28576410e-01
2.71208018e-01 -1.53250885e+00 1.32891405e+00 6.54234707e-01
-7.60043204e-01 -3.86448324e-01 -9.96515810e-01 1.89479083e-01
-1.72040071e-02 9.31379199e-01 -4.00837719e-01 2.44333133e-01
7.14640796e-01 8.09998214e-02 -2.44695067e-01 7.56274343e-01
-5.58271885e-01 7.59423494e-01 -3.61675113e-01 2.27223206e-02
1.26156643e-01 3.36626172e-01 8.34694028e-01 1.03678644e+00
2.64680743e-01 1.50375232e-01 -3.07310149e-02 7.82124043e-01
-3.89327079e-01 -4.36202765e-01 -6.44168973e-01 4.26760793e-01
1.04595554e+00 8.70633423e-01 -1.10478409e-01 -3.02295834e-01
-1.49848475e-03 1.01426613e+00 2.83181399e-01 2.58781433e-01
-1.22474861e+00 -7.12297201e-01 1.06531799e+00 -5.99407315e-01
1.48436427e-01 2.77794659e-01 -7.21917003e-02 -9.53800917e-01
-1.69113949e-01 -6.06377602e-01 4.85672802e-01 -6.61430895e-01
-1.98218286e+00 8.05077553e-01 1.79103404e-01 -9.05789793e-01
-2.53862768e-01 -6.84272051e-01 -1.18971455e+00 1.00335586e+00
-1.19874108e+00 -8.94610345e-01 -2.15144884e-02 1.07258451e+00
2.22005740e-01 -4.81858999e-01 1.00012732e+00 1.90103054e-01
-8.74435484e-01 1.28622067e+00 2.92940319e-01 4.44899857e-01
7.59170949e-01 -1.29834545e+00 4.41648364e-01 8.36893797e-01
-4.01668817e-01 3.42164606e-01 9.60182488e-01 -8.14764678e-01
-6.35957778e-01 -1.47199535e+00 1.33049190e-01 -7.32350826e-01
7.78376579e-01 -4.26099658e-01 -1.08096921e+00 6.93656802e-01
3.40837315e-02 5.61267510e-02 9.95525599e-01 -2.56768316e-01
-5.72504222e-01 5.42642251e-02 -1.81150627e+00 6.51653349e-01
7.53324330e-01 -5.59604943e-01 -1.00109851e+00 2.88892627e-01
1.09320486e+00 -5.51461935e-01 -9.20443892e-01 5.25617719e-01
2.64287651e-01 -8.62256587e-01 1.19281459e+00 -8.39966536e-01
3.66515070e-01 1.73906218e-02 -5.49280941e-01 -1.42741060e+00
-4.01182801e-01 -4.79764819e-01 -6.71404958e-01 1.72927666e+00
1.67952001e-01 -6.94603562e-01 5.65080285e-01 8.99738312e-01
-2.12415665e-01 -5.58460712e-01 -1.18237448e+00 -5.32380998e-01
5.41471183e-01 -7.49229312e-01 9.31685925e-01 7.87972987e-01
-3.21597338e-01 -4.41322513e-02 -4.75669414e-01 9.38291669e-01
1.12726140e+00 -7.12242544e-01 2.79158413e-01 -1.41796744e+00
-2.14593276e-01 -1.66830242e-01 -8.51542950e-01 -4.34284508e-01
5.88846982e-01 -3.73618156e-01 1.45044848e-01 -9.20402646e-01
-1.48336336e-01 -3.22957523e-04 -6.57164574e-01 2.97821164e-01
-5.23645639e-01 -4.57040183e-02 -1.65466487e-01 -2.37655230e-02
-5.19726694e-01 7.92730570e-01 5.11139691e-01 -1.05037622e-01
2.94428498e-01 3.94190073e-01 -8.98151159e-01 1.43037605e+00
1.04953206e+00 -6.76096737e-01 -5.06342709e-01 -3.53513718e-01
-1.03805913e-02 -1.28528327e-01 2.67385364e-01 -1.34982097e+00
3.14721078e-01 1.22968197e-01 8.77533257e-01 -5.35570860e-01
2.85204768e-01 -8.33407998e-01 -2.58000970e-01 1.57959566e-01
-3.38283628e-01 -3.47455531e-01 3.68540049e-01 1.15423048e+00
7.68413618e-02 -6.38417065e-01 9.05608952e-01 -1.39411360e-01
-4.27911818e-01 5.83538711e-01 -3.80847573e-01 2.61900514e-01
1.22643626e+00 4.44477260e-01 -6.40423179e-01 -4.83255088e-01
-1.04917598e+00 1.90407261e-01 -1.28548309e-01 2.56536186e-01
7.78759241e-01 -1.61410284e+00 -7.67450392e-01 4.50665168e-02
-1.41576156e-01 2.96441972e-01 4.93904561e-01 2.10280344e-02
3.49555351e-02 -1.55689344e-01 -2.77954251e-01 -1.41834959e-01
-8.91105115e-01 7.42891252e-01 3.66537601e-01 -1.35025650e-01
-2.55323917e-01 1.48255336e+00 1.69579178e-01 -5.11859655e-01
8.61795962e-01 2.31197223e-01 -3.34667921e-01 3.80050428e-02
9.83060777e-01 4.53168988e-01 -2.82160550e-01 -4.74442810e-01
-3.51639032e-01 2.26472039e-03 -4.45460737e-01 -2.29111668e-02
9.91953373e-01 -4.45108153e-02 1.73402265e-01 2.21611708e-01
9.58429754e-01 -1.94806322e-01 -1.37285590e+00 -5.95882118e-01
-4.13955033e-01 -2.30432332e-01 8.23454261e-02 -7.82531083e-01
-9.71292078e-01 1.05604446e+00 6.28566623e-01 2.17793584e-01
1.07928157e+00 -1.43701341e-02 9.35009897e-01 3.85390341e-01
2.62934655e-01 -8.46843421e-01 -1.79501593e-01 7.18572617e-01
8.63399267e-01 -1.07509768e+00 -2.66407669e-01 1.29295051e-01
-7.97453642e-01 9.66554344e-01 8.52362037e-01 -2.83136129e-01
1.06380689e+00 6.38725579e-01 -5.53539507e-02 2.42384329e-01
-5.98827004e-01 1.83317110e-01 6.07022978e-02 1.10025883e+00
-3.13027769e-01 9.39912945e-02 4.34907734e-01 1.49817681e+00
-3.14718574e-01 -5.79726517e-01 4.46699321e-01 6.02688670e-01
-7.99834967e-01 -7.61980593e-01 -7.31627524e-01 4.32862937e-01
-7.76665092e-01 -2.03429386e-01 -5.77358603e-01 1.76449448e-01
2.36781240e-01 1.00826299e+00 -1.34140775e-01 -8.44982266e-01
1.90314546e-01 2.69770056e-01 2.20228508e-01 -3.45988750e-01
-2.83715844e-01 -4.77395326e-01 -3.14837575e-01 -3.28182817e-01
2.58268923e-01 -5.41779935e-01 -1.22243178e+00 -5.35656333e-01
-2.20458582e-01 1.06962398e-02 5.32729924e-01 1.11407185e+00
1.91049203e-01 7.89346099e-01 8.58814538e-01 -9.26789165e-01
-1.26344633e+00 -1.17791772e+00 -8.16181660e-01 1.74526483e-01
1.85962394e-01 -8.35087061e-01 -9.61844385e-01 -4.21540141e-01] | [5.615905284881592, 7.8759589195251465] |
e794b42a-15b9-452a-91b6-84781c7d44c9 | variational-graph-recurrent-neural-networks | 1908.09710 | null | https://arxiv.org/abs/1908.09710v3 | https://arxiv.org/pdf/1908.09710v3.pdf | Variational Graph Recurrent Neural Networks | Representation learning over graph structured data has been mostly studied in static graph settings while efforts for modeling dynamic graphs are still scant. In this paper, we develop a novel hierarchical variational model that introduces additional latent random variables to jointly model the hidden states of a graph recurrent neural network (GRNN) to capture both topology and node attribute changes in dynamic graphs. We argue that the use of high-level latent random variables in this variational GRNN (VGRNN) can better capture potential variability observed in dynamic graphs as well as the uncertainty of node latent representation. With semi-implicit variational inference developed for this new VGRNN architecture (SI-VGRNN), we show that flexible non-Gaussian latent representations can further help dynamic graph analytic tasks. Our experiments with multiple real-world dynamic graph datasets demonstrate that SI-VGRNN and VGRNN consistently outperform the existing baseline and state-of-the-art methods by a significant margin in dynamic link prediction. | ['Krishna R. Narayanan', 'Xiaoning Qian', 'Ehsan Hajiramezanali', 'Arman Hasanzadeh', 'Nick Duffield', 'Mingyuan Zhou'] | 2019-08-26 | variational-graph-recurrent-neural-networks-1 | http://papers.nips.cc/paper/9254-variational-graph-recurrent-neural-networks | http://papers.nips.cc/paper/9254-variational-graph-recurrent-neural-networks.pdf | neurips-2019-12 | ['dynamic-link-prediction'] | ['graphs'] | [ 2.14690920e-02 7.13122845e-01 -6.71451330e-01 -9.57919136e-02
-3.22820753e-01 -3.98390412e-01 8.05648208e-01 -1.29636645e-01
6.67702496e-01 5.07561982e-01 4.22038168e-01 -4.33728188e-01
-2.17402279e-01 -1.00690091e+00 -6.66532040e-01 -4.95588213e-01
-6.00217223e-01 9.52102184e-01 8.88753235e-02 -1.88761756e-01
-4.01266217e-01 2.80915350e-01 -8.97901595e-01 -1.84027046e-01
6.21381938e-01 4.39666450e-01 -3.32807377e-02 7.91840136e-01
-7.99854323e-02 1.12805831e+00 -2.20996305e-01 -4.40373868e-01
6.94907680e-02 -9.71823037e-02 -6.28769100e-01 1.13748871e-01
1.04473695e-01 -2.10771054e-01 -1.10963881e+00 1.00686502e+00
7.72697404e-02 3.28054905e-01 9.22181129e-01 -1.51732600e+00
-1.08546007e+00 1.41984797e+00 -5.85661650e-01 2.40770042e-01
-1.50173232e-01 -1.24668586e-04 1.60135663e+00 -1.48086086e-01
1.07568085e+00 1.65555620e+00 8.33552122e-01 5.08028746e-01
-1.87472999e+00 -3.60550791e-01 8.26098084e-01 -1.36515684e-03
-1.33414245e+00 -1.89055771e-01 1.23868191e+00 -4.70452845e-01
1.14583147e+00 -4.60104644e-02 6.20887220e-01 1.70994973e+00
4.46353048e-01 8.01463246e-01 3.89579564e-01 5.56104183e-02
-1.47698998e-01 -3.64620715e-01 3.68970156e-01 9.53060687e-01
3.79043430e-01 1.47695690e-01 -3.73991817e-01 -4.29227203e-01
1.03075886e+00 2.94286638e-01 -6.94501773e-02 -7.67650902e-01
-6.11377358e-01 1.10637641e+00 6.93986237e-01 1.91113487e-01
-5.40083110e-01 1.05049598e+00 4.61314678e-01 2.49403641e-01
8.44910264e-01 -7.90068060e-02 -3.41049701e-01 -1.34094670e-01
-9.31504369e-01 5.41510694e-02 8.52124751e-01 1.01964509e+00
6.86566353e-01 6.98671937e-01 -2.79228002e-01 6.46639705e-01
8.41097832e-01 3.86684299e-01 -3.72163244e-02 -7.90171623e-01
3.31296265e-01 7.12605536e-01 -4.35672164e-01 -1.36853492e+00
-2.89182395e-01 -6.32416964e-01 -1.04306257e+00 -3.92096221e-01
-2.53425352e-02 6.11396134e-02 -1.36137390e+00 2.13136816e+00
-2.47072987e-02 5.94096184e-01 -1.57039851e-01 1.99074730e-01
9.43668604e-01 8.26527834e-01 6.84162304e-02 -2.31299952e-01
5.95507562e-01 -8.44476342e-01 -8.08513641e-01 -8.41182470e-02
4.40557063e-01 2.03220919e-01 7.19958663e-01 -1.36908978e-01
-9.59939301e-01 -2.36504883e-01 -6.83351696e-01 -9.67816822e-03
-1.86207280e-01 -4.97376651e-01 1.19094813e+00 4.13742989e-01
-1.56338799e+00 7.09452510e-01 -1.54395914e+00 -3.19234371e-01
1.88933969e-01 2.37145290e-01 1.19737359e-02 1.14422373e-01
-1.28380370e+00 3.05123538e-01 1.52091801e-01 3.85372579e-01
-1.46740735e+00 -6.31738365e-01 -1.09324133e+00 3.78698230e-01
6.86493516e-01 -8.48960102e-01 8.53308022e-01 -4.55458879e-01
-1.46811628e+00 4.98941004e-01 -2.70653188e-01 -7.83711791e-01
3.29743743e-01 2.01334238e-01 -5.07668138e-01 -8.71469453e-03
-8.02818313e-02 3.20765793e-01 8.82004559e-01 -1.38971269e+00
4.45280880e-01 -2.53260851e-01 -1.66171908e-01 -1.24203719e-01
-5.52196465e-02 -4.74562585e-01 -7.44760692e-01 -6.94671154e-01
2.21820965e-01 -1.17789686e+00 -4.69250470e-01 -4.15980250e-01
-8.22478771e-01 -4.89754856e-01 9.23370779e-01 -6.56061769e-01
1.52984059e+00 -1.65367162e+00 6.53218031e-01 5.72455049e-01
9.66543674e-01 -9.95301530e-02 -3.14498872e-01 5.98457992e-01
3.20341401e-02 4.90070879e-01 1.39963225e-01 -7.13200033e-01
3.19483668e-01 6.65004253e-01 -6.33662343e-01 2.67367333e-01
-1.57251954e-01 1.66897941e+00 -8.89923632e-01 -3.66674781e-01
2.20901016e-02 7.45824099e-01 -5.71825087e-01 -3.89612801e-02
-7.50212550e-01 1.20121203e-01 -4.88845438e-01 7.61765957e-01
3.29814643e-01 -1.06559920e+00 8.54406774e-01 4.58745323e-02
7.05663264e-01 9.85910371e-02 -8.43098104e-01 1.44155645e+00
-1.34452432e-01 8.01547527e-01 2.56795324e-02 -9.11905169e-01
8.84479463e-01 1.26541287e-01 5.91430068e-01 -2.87182242e-01
-1.96387302e-02 -3.69591147e-01 3.86330253e-03 9.26178545e-02
5.46252012e-01 3.37169737e-01 -1.07305713e-01 5.65595269e-01
2.31137529e-01 4.13047314e-01 1.56143293e-01 1.04164028e+00
1.39983153e+00 2.02255964e-01 -2.79900096e-02 -1.03747532e-01
-9.55577642e-02 -5.29176414e-01 7.41048098e-01 1.04505241e+00
-2.61596531e-01 1.69207945e-01 1.07883668e+00 -3.16432744e-01
-9.70004737e-01 -1.45308793e+00 3.47457647e-01 1.14181566e+00
-1.57155901e-01 -9.16495025e-01 -2.82379866e-01 -6.87353194e-01
3.54779929e-01 5.80254078e-01 -9.99255896e-01 -3.31814289e-01
-4.75629449e-01 -7.65540481e-01 3.62854779e-01 6.37127101e-01
-2.25296706e-01 -9.73391712e-01 4.06988233e-01 4.65790957e-01
5.85364848e-02 -1.12727273e+00 -3.16061527e-01 5.19787632e-02
-1.00792050e+00 -8.58947933e-01 -2.16479689e-01 -3.14437717e-01
5.58926463e-01 1.43734157e-01 1.58672631e+00 1.83574557e-01
-1.07400857e-01 5.98637581e-01 -2.45119676e-01 1.59269065e-01
-7.24550605e-01 5.86228609e-01 2.28200015e-02 -1.80914670e-01
-1.29322875e-02 -1.15567183e+00 -2.17520222e-01 -1.00234628e-01
-6.86494470e-01 2.58079112e-01 3.33399206e-01 7.57442713e-01
5.62148690e-01 -2.32680310e-02 2.86519289e-01 -1.51328194e+00
6.85908496e-01 -8.53029966e-01 -6.57721698e-01 3.37078393e-01
-1.27735436e+00 5.32849312e-01 3.72698337e-01 -6.55937791e-01
-6.61229134e-01 -6.46252453e-01 3.33402425e-01 -9.65118647e-01
4.43445623e-01 1.11024725e+00 5.10252826e-02 1.87234879e-01
2.51276881e-01 1.45279303e-01 6.34845197e-02 -2.96212643e-01
7.57179201e-01 -2.34367013e-01 3.09713155e-01 -3.80854160e-01
1.13077939e+00 4.22312886e-01 4.10979331e-01 -4.99818683e-01
-5.51450431e-01 -1.14025347e-01 -4.00120646e-01 -2.57951587e-01
5.93763411e-01 -9.71492231e-01 -6.38929784e-01 3.62608373e-01
-7.39217460e-01 -8.99019182e-01 -1.82871357e-01 -7.04076588e-02
-5.02652049e-01 5.05189598e-01 -1.12443066e+00 -9.88631904e-01
-3.86067986e-01 -1.00628316e+00 1.14420378e+00 -6.47609681e-02
-3.78856696e-02 -1.95062149e+00 2.63442606e-01 -1.96261238e-02
5.15330017e-01 6.32058978e-01 1.17192221e+00 -5.20845532e-01
-1.06471205e+00 -1.22949027e-01 -2.11756691e-01 -2.82628596e-01
9.73601043e-02 4.37853783e-01 -3.01516324e-01 -6.00124836e-01
-7.74323344e-01 -9.71279368e-02 1.21158624e+00 8.27646196e-01
9.13305938e-01 -4.04214889e-01 -7.84274042e-01 1.01192224e+00
1.34248543e+00 -4.44859356e-01 4.96634662e-01 -3.47400963e-01
1.43318117e+00 1.31405696e-01 -1.62730590e-01 2.19699472e-01
8.70653212e-01 6.21978641e-01 7.66469479e-01 4.57028374e-02
-1.82173014e-01 -8.18304777e-01 5.19186497e-01 1.07755315e+00
-1.14971645e-01 -7.71576881e-01 -1.03651869e+00 5.53172648e-01
-2.22209120e+00 -1.00054157e+00 -3.25503141e-01 1.39113474e+00
3.20845664e-01 3.48506212e-01 7.57017061e-02 -5.57409585e-01
8.65623474e-01 9.62202430e-01 -9.17402744e-01 -1.98607564e-01
-9.38059241e-02 -1.43483713e-01 5.96618414e-01 8.04738581e-01
-8.05844903e-01 1.35236526e+00 6.83026600e+00 5.60160220e-01
-7.06909955e-01 8.66236985e-02 4.43035156e-01 7.82501101e-02
-9.31567073e-01 3.67116183e-01 -7.78703570e-01 2.70417660e-01
1.41704869e+00 -3.20151001e-01 7.75501490e-01 7.66406178e-01
7.70130828e-02 6.74201488e-01 -9.60630357e-01 7.68159270e-01
-1.05293758e-01 -1.65277898e+00 5.05192757e-01 6.77552760e-01
1.01006019e+00 5.47039092e-01 2.99863160e-01 6.85254574e-01
1.35345352e+00 -1.21482301e+00 1.40526608e-01 9.71343696e-01
6.58836544e-01 -6.03272974e-01 3.40244263e-01 1.11514308e-01
-1.62346482e+00 1.16868183e-01 -3.31753641e-01 3.72948706e-01
3.77115846e-01 3.46296668e-01 -8.45747590e-01 7.40046263e-01
3.18835348e-01 1.31898379e+00 -6.64751530e-01 1.35313138e-01
-3.21599156e-01 1.24033403e+00 -2.58360118e-01 1.64879784e-01
2.83367395e-01 -3.75757605e-01 1.01881707e+00 7.58036077e-01
-3.35682333e-02 -3.77248973e-01 4.56564218e-01 1.41315567e+00
-3.19143087e-01 -2.40538076e-01 -7.75546491e-01 -8.39507759e-01
4.35615689e-01 1.09703279e+00 -7.59423554e-01 -2.23337680e-01
-3.69239524e-02 7.83262849e-01 8.26825798e-01 9.18538928e-01
-8.94826591e-01 2.20330611e-01 7.56233215e-01 1.60163671e-01
4.33501095e-01 -5.72026432e-01 3.05624247e-01 -1.53042257e+00
-2.93470800e-01 -4.46398288e-01 5.99628150e-01 -6.28069937e-01
-1.43524837e+00 6.42762303e-01 6.85358271e-02 -5.52514076e-01
-9.75499451e-01 -3.77241343e-01 -7.07875133e-01 8.04358184e-01
-1.17602062e+00 -1.70722365e+00 -1.36506781e-01 5.41035116e-01
2.58450955e-01 -1.64142489e-01 6.74150884e-01 -2.51172990e-01
-7.95907617e-01 6.30943716e-01 2.89479285e-01 3.08418036e-01
-1.73909497e-02 -1.50008345e+00 1.07208920e+00 1.07011151e+00
5.62437534e-01 8.58517706e-01 8.88444722e-01 -1.24693954e+00
-1.60803223e+00 -1.38281059e+00 4.27585244e-01 -6.23105586e-01
1.06324553e+00 -7.24651873e-01 -1.10270166e+00 1.45010126e+00
-1.41978920e-01 2.78280318e-01 5.83776295e-01 7.04307854e-01
-4.63549525e-01 3.26349616e-01 -6.13162398e-01 6.55840099e-01
1.43906951e+00 -9.03108120e-01 5.09033948e-02 3.30268085e-01
1.28190148e+00 -3.08969647e-01 -1.09435773e+00 4.03569460e-01
4.19257402e-01 -4.38273668e-01 9.55953240e-01 -1.00506926e+00
1.08563051e-01 3.29548903e-02 -1.21135928e-01 -1.27635586e+00
-8.18777442e-01 -1.19359148e+00 -1.27209008e+00 1.28237855e+00
3.90685767e-01 -6.44125164e-01 1.17183805e+00 5.11514783e-01
5.76381311e-02 -5.47289491e-01 -5.61921418e-01 -5.95637321e-01
-3.18952054e-02 -4.22974408e-01 4.61887151e-01 8.73398662e-01
-4.14787948e-01 3.52141738e-01 -6.75471842e-01 4.01432753e-01
8.51345539e-01 1.58098236e-01 9.30653572e-01 -1.68539059e+00
-5.47564268e-01 -4.55972880e-01 -7.50287414e-01 -1.09182382e+00
8.72214675e-01 -1.39311159e+00 -4.55769449e-01 -1.90047371e+00
4.47696060e-01 -1.02321021e-01 -3.67999762e-01 3.46312314e-01
-3.75072747e-01 -1.81628525e-01 1.60076562e-02 3.75815302e-01
-8.75675619e-01 1.02719772e+00 8.57399702e-01 -1.93324655e-01
-3.41956705e-01 -2.20993925e-02 -6.70099616e-01 3.85825247e-01
3.58568192e-01 -4.05862689e-01 -8.74820948e-01 -1.29465461e-01
4.39533472e-01 3.16737175e-01 4.54115301e-01 -3.80626827e-01
1.62819773e-01 -1.12888992e-01 5.77423424e-02 -8.24465513e-01
1.90646097e-01 -4.27951396e-01 6.28427207e-01 2.96202093e-01
-4.18476582e-01 1.77631617e-01 -2.94238746e-01 1.43291414e+00
8.92660990e-02 4.95654643e-01 2.41989329e-01 -2.02840231e-02
-6.09879851e-01 1.01889038e+00 -2.86078781e-01 1.13608718e-01
6.28116786e-01 -1.73697993e-02 -4.23786640e-01 -1.00309908e+00
-1.19956172e+00 6.25601113e-01 4.45772171e-01 6.22093797e-01
4.96244431e-01 -1.34669030e+00 -5.73018849e-01 -1.10108510e-03
-6.26053289e-02 -1.27627999e-01 4.65543151e-01 5.74024796e-01
-8.76648948e-02 2.12026730e-01 2.98997223e-01 -7.17682600e-01
-1.03522074e+00 6.76132441e-01 3.37549299e-01 -1.22380221e+00
-1.00935876e+00 7.16003180e-01 -2.78557390e-02 -5.61608016e-01
1.17161468e-01 -3.67161810e-01 -1.25096411e-01 -8.58101770e-02
-2.88177103e-01 2.74100721e-01 -6.69190228e-01 -7.65914738e-01
-1.73734166e-02 3.28940563e-02 -3.18371594e-01 9.31962878e-02
1.52874660e+00 -3.83799732e-01 -2.04952255e-01 8.08040440e-01
1.15490627e+00 -1.09581284e-01 -1.54169405e+00 -5.15109420e-01
6.01217896e-02 5.89711070e-02 1.52484536e-01 -2.97688007e-01
-1.27144921e+00 6.91652417e-01 1.41883001e-01 5.50470293e-01
5.16644120e-01 3.83719593e-01 5.38938105e-01 2.78395951e-01
4.41768855e-01 -6.78167701e-01 1.25873491e-01 6.34753466e-01
6.84356868e-01 -9.63267088e-01 6.22753613e-02 -6.03304863e-01
-4.06722158e-01 8.95833015e-01 5.11325598e-01 -4.83115226e-01
9.71672535e-01 8.96361656e-03 -4.00494009e-01 -5.94808459e-01
-1.32632303e+00 -1.56901479e-01 4.44362551e-01 6.77694142e-01
1.97382629e-01 3.43601227e-01 4.93841529e-01 4.50263202e-01
-1.71833143e-01 -4.63980615e-01 6.09680772e-01 5.02927601e-01
5.66997416e-02 -1.17797232e+00 4.54693317e-01 6.73944175e-01
-3.28001827e-01 -2.36280680e-01 -3.89285922e-01 7.81419873e-01
-7.82159925e-01 6.84570253e-01 1.14063889e-01 -4.88376915e-01
-2.53731191e-01 -2.17961241e-02 2.40100101e-01 -7.68187881e-01
-1.92012683e-01 1.83159143e-01 1.31945461e-01 -7.35639691e-01
-5.42543940e-02 -7.50577807e-01 -1.07979417e+00 -5.01226842e-01
-2.58333921e-01 1.00492582e-01 2.53933251e-01 5.83995044e-01
6.27733171e-01 1.02313673e+00 3.57019514e-01 -7.55793512e-01
-5.18623948e-01 -1.09967411e+00 -8.64838958e-01 3.57018679e-01
3.38980496e-01 -8.46343040e-01 -5.66269457e-01 -2.51811177e-01] | [7.143710136413574, 6.056694507598877] |
45a42885-e58a-4727-9834-1a007ef0c56d | structured-learning-in-time-dependent-cox | 2306.12528 | null | https://arxiv.org/abs/2306.12528v1 | https://arxiv.org/pdf/2306.12528v1.pdf | Structured Learning in Time-dependent Cox Models | Cox models with time-dependent coefficients and covariates are widely used in survival analysis. In high-dimensional settings, sparse regularization techniques are employed for variable selection, but existing methods for time-dependent Cox models lack flexibility in enforcing specific sparsity patterns (i.e., covariate structures). We propose a flexible framework for variable selection in time-dependent Cox models, accommodating complex selection rules. Our method can adapt to arbitrary grouping structures, including interaction selection, temporal, spatial, tree, and directed acyclic graph structures. It achieves accurate estimation with low false alarm rates. We develop the sox package, implementing a network flow algorithm for efficiently solving models with complex covariate structures. Sox offers a user-friendly interface for specifying grouping structures and delivers fast computation. Through examples, including a case study on identifying predictors of time to all-cause death in atrial fibrillation patients, we demonstrate the practical application of our method with specific selection rules. | ['Mireille E. Schnitzer', 'Marc Dorais', 'Sylvie Perreault', 'Rui Wang', 'Robert W. Platt', 'Archer Y. Yang', 'Yi Lian', 'Guanbo Wang'] | 2023-06-21 | null | null | null | null | ['variable-selection', 'survival-analysis'] | ['methodology', 'miscellaneous'] | [ 8.87499526e-02 -5.42865515e-01 -7.02900589e-01 -5.12578309e-01
-7.88647413e-01 -3.18849236e-01 -1.47776222e-02 3.51876825e-01
-2.29502097e-01 9.91041243e-01 3.72844458e-01 -8.95109713e-01
-5.99226415e-01 -7.67755210e-01 -9.99028832e-02 -6.00203097e-01
-1.11841571e+00 6.33693695e-01 -7.09666014e-02 8.83020386e-02
5.85760036e-03 5.85129082e-01 -8.79357815e-01 -8.53734016e-02
6.59184396e-01 4.14728492e-01 -2.64320046e-01 6.48351729e-01
2.61654168e-01 4.68799055e-01 -2.21611515e-01 4.52386886e-02
1.02496542e-01 -5.76202750e-01 -4.01745379e-01 1.01451362e-02
-8.57193768e-02 -9.64847878e-02 -2.78991610e-01 2.84488380e-01
7.24396229e-01 -5.38862236e-02 8.10066938e-01 -1.20789754e+00
9.43727121e-02 5.45640647e-01 -6.01661325e-01 4.03537512e-01
-1.54727967e-02 7.09969625e-02 7.13741839e-01 -1.03322756e+00
4.37966794e-01 1.12423968e+00 1.21638894e+00 3.24353307e-01
-1.71988416e+00 -5.91817141e-01 1.37784988e-01 -1.46687612e-01
-1.63517082e+00 -5.67284703e-01 4.02873725e-01 -6.23208046e-01
6.97342992e-01 6.97446465e-01 7.18586981e-01 8.47116292e-01
5.73589325e-01 3.26072216e-01 8.86216938e-01 -1.63793489e-01
4.67837006e-01 -3.11642647e-01 5.34707844e-01 7.23331034e-01
3.65771443e-01 4.64540541e-01 -4.22341585e-01 -1.36968768e+00
1.15562952e+00 2.35553578e-01 -6.81676120e-02 -6.38868511e-01
-1.29146659e+00 1.20965016e+00 -2.34370962e-01 -9.05574039e-02
-3.80355150e-01 2.17350766e-01 6.95778489e-01 2.98931599e-01
4.21646774e-01 3.82843763e-02 -5.69636881e-01 1.82681814e-01
-1.04812813e+00 1.94792062e-01 8.60369921e-01 1.06080329e+00
2.94431895e-01 1.79073155e-01 -5.59765100e-01 8.27363670e-01
2.89569139e-01 2.64786988e-01 -1.13223074e-02 -8.88848305e-01
1.05515607e-01 1.67600051e-01 5.45075424e-02 -6.86547399e-01
-1.16700029e+00 -8.32636356e-01 -1.51061642e+00 5.52361039e-03
5.00669122e-01 -4.90968496e-01 -8.46401930e-01 1.80032086e+00
4.95861888e-01 3.63555044e-01 -3.00859213e-01 6.51003957e-01
4.93247896e-01 1.19372733e-01 4.82038617e-01 -1.01879168e+00
1.51706862e+00 -3.82950634e-01 -6.67170048e-01 6.13817573e-02
1.22649491e+00 -2.94939548e-01 6.46220982e-01 3.15063655e-01
-9.50053573e-01 1.76655516e-01 -5.49091578e-01 1.38146222e-01
1.73946485e-01 -9.03712586e-02 1.10749936e+00 6.17505491e-01
-9.41382408e-01 5.34863412e-01 -1.13363552e+00 -2.53809661e-01
3.60073388e-01 5.06939590e-01 -2.03790471e-01 -5.01106121e-02
-1.20165408e+00 5.58899760e-01 -1.04303516e-01 1.86138581e-02
-9.64831173e-01 -1.09647167e+00 -9.07648981e-01 1.49779603e-01
3.05747777e-01 -1.44811010e+00 7.93182552e-01 -3.84639502e-01
-8.88752818e-01 6.53654218e-01 -5.27072191e-01 -4.13402259e-01
6.44240022e-01 3.08787704e-01 -1.84681937e-01 -2.49785841e-01
1.92511961e-01 -1.74177974e-01 7.38331199e-01 -6.42656684e-01
-3.55187148e-01 -4.11686122e-01 -3.95112067e-01 6.78722262e-02
-1.71083406e-01 4.22197312e-01 -2.67139584e-01 -9.90443528e-01
1.88765779e-01 -8.59801114e-01 -1.17323518e+00 3.02736700e-01
-3.93060118e-01 9.05832276e-02 3.58332992e-01 -4.42143083e-01
1.86726356e+00 -2.05133390e+00 4.40139920e-02 7.02018261e-01
3.47283036e-01 -6.26996577e-01 1.35633290e-01 4.85261232e-01
-3.55622083e-01 2.71376491e-01 -7.43072689e-01 -4.72258538e-01
-4.46755588e-01 3.04550409e-01 3.36651713e-01 7.96557605e-01
-1.86087996e-01 5.25664866e-01 -8.58029962e-01 -7.61119902e-01
5.88560663e-03 4.70519543e-01 -7.66399264e-01 -1.25913531e-01
2.95843959e-01 5.83696842e-01 -7.46911347e-01 6.82682812e-01
4.69432086e-01 -6.77097201e-01 6.28574610e-01 2.93206185e-01
-2.08364189e-01 -7.41125047e-02 -1.23042083e+00 1.52248132e+00
-3.48481029e-01 1.14593089e-01 4.75349635e-01 -1.03218150e+00
8.18381250e-01 5.47374249e-01 9.20514286e-01 8.01875368e-02
-9.72321630e-02 1.84605315e-01 -2.06957489e-01 -2.43205354e-01
-2.53012300e-01 -5.25651097e-01 -1.90590918e-01 3.32540631e-01
-3.87800336e-01 1.40431717e-01 2.16646120e-01 2.57442415e-01
1.49031007e+00 -6.18609190e-01 8.32669079e-01 -6.02754533e-01
2.79954404e-01 1.37856469e-01 9.68271911e-01 9.64629650e-01
2.23796461e-02 6.65618062e-01 8.29866469e-01 -8.09798241e-01
-8.27854931e-01 -9.18465912e-01 -7.53972590e-01 7.51277149e-01
-5.15983224e-01 -6.24493957e-01 1.87686682e-01 -4.70818639e-01
2.64346540e-01 3.85806948e-01 -6.51613474e-01 -1.48007961e-03
-7.59940326e-01 -1.43779182e+00 3.90468359e-01 4.96615320e-01
-2.39723250e-01 -5.20218730e-01 -3.44157815e-01 5.88742912e-01
1.49192363e-01 -4.06477004e-01 -7.06177473e-01 6.15640938e-01
-1.54092133e+00 -1.22631919e+00 -6.89521313e-01 -7.96457529e-01
8.54023635e-01 3.37690152e-02 1.22998607e+00 3.32262307e-01
-4.15633291e-01 2.48358041e-01 2.04021558e-01 -1.13417663e-01
-6.85415268e-02 -9.77214128e-02 1.22320931e-02 -3.18956167e-01
-9.77425799e-02 -5.89031041e-01 -8.81927431e-01 4.84995306e-01
-6.64055049e-01 -1.44318387e-01 8.30135122e-02 1.35245013e+00
8.22066545e-01 -1.57621488e-01 6.36425853e-01 -1.28920901e+00
6.77009046e-01 -1.03272891e+00 -5.89119196e-01 -4.05539759e-03
-9.45284784e-01 -2.39651427e-01 3.99414212e-01 -4.14894581e-01
-6.10233068e-01 4.42406446e-01 1.25670880e-02 -2.38415882e-01
1.34940103e-01 1.21186578e+00 1.86723128e-01 1.79521944e-02
6.15302444e-01 -1.57034714e-02 8.67255703e-02 -4.01877791e-01
-7.03783557e-02 5.14197424e-02 2.51287352e-02 -5.10052264e-01
4.20861214e-01 5.67532063e-01 6.74537957e-01 -6.54809415e-01
-2.57580996e-01 -4.71149206e-01 -5.24851680e-01 2.89361596e-01
4.48522389e-01 -1.02936196e+00 -5.83611965e-01 2.36928120e-01
-7.09720731e-01 -3.69421065e-01 -1.45038456e-01 8.11837733e-01
-6.00890040e-01 3.24161559e-01 -6.93361521e-01 -5.82935572e-01
-3.41296285e-01 -9.10738349e-01 8.07657659e-01 -4.71959889e-01
-5.71911156e-01 -1.47827375e+00 2.17271686e-01 -4.62157071e-01
4.83434409e-01 4.20863420e-01 1.33045793e+00 -5.66027164e-01
-3.15688461e-01 -1.74072310e-01 1.14824563e-01 -3.68241668e-01
2.37328008e-01 -5.69499731e-02 -1.83329865e-01 -6.16269708e-01
4.16227393e-02 2.74522930e-01 6.71469629e-01 1.27206206e+00
1.57627416e+00 -4.33662534e-01 -8.77892494e-01 1.04567540e+00
1.16322708e+00 1.76401779e-01 4.95322734e-01 6.21391237e-02
4.96126026e-01 5.17697573e-01 5.19325376e-01 1.13802600e+00
3.72317731e-01 6.71800435e-01 1.22431815e-01 -7.10993886e-01
4.44002032e-01 2.52898961e-01 -1.93626478e-01 2.98188031e-01
9.57484096e-02 -1.12443320e-01 -1.06189489e+00 5.20959020e-01
-1.93955696e+00 -7.14460731e-01 -7.20278978e-01 2.47323298e+00
9.82464373e-01 -1.20954812e-01 3.84420067e-01 2.20565032e-02
6.74162865e-01 -8.14315304e-03 -5.46759725e-01 -2.73789197e-01
-3.22089158e-02 2.34350234e-01 7.46940434e-01 6.66634321e-01
-1.22680879e+00 3.77558500e-01 8.05867577e+00 6.49286747e-01
-7.14792252e-01 1.54855609e-01 8.37706804e-01 -1.87097117e-01
-3.37982088e-01 2.23264039e-01 -6.20987296e-01 2.03688189e-01
9.78422821e-01 -5.08490562e-01 -1.37738645e-01 4.72471625e-01
1.02380383e+00 2.13061064e-01 -9.28021252e-01 7.72957683e-01
-5.37699759e-01 -1.45467925e+00 -4.41130400e-01 3.74656320e-02
4.19339508e-01 -3.00754815e-01 -2.61632860e-01 5.92317060e-02
5.19476235e-01 -1.05078650e+00 -4.89792936e-02 2.70086527e-01
1.14729059e+00 -5.88483334e-01 4.62387085e-01 6.84466660e-02
-1.16843688e+00 -9.89325792e-02 8.34633037e-02 -1.17832102e-01
7.47676730e-01 1.09777343e+00 -6.66230261e-01 2.90561199e-01
4.00968611e-01 1.07784534e+00 -7.94625133e-02 1.46429443e+00
1.67571813e-01 8.83624554e-01 -5.25021493e-01 4.35074508e-01
-2.21075684e-01 -2.04543740e-01 6.68742895e-01 1.23047316e+00
5.05265892e-01 4.64230627e-01 3.70490700e-01 3.92651796e-01
4.91203606e-01 2.52032250e-01 -6.41677380e-01 4.68645304e-01
6.26481056e-01 8.41028929e-01 -7.76617944e-01 -3.43749464e-01
-6.18617177e-01 3.50616217e-01 -1.17808498e-01 6.03437722e-01
-7.34433591e-01 -8.45694635e-03 7.17812419e-01 5.64536631e-01
3.09851561e-02 -5.30530632e-01 -8.38076830e-01 -1.06342173e+00
-3.42708826e-01 -8.74914885e-01 1.32781613e+00 -1.40948310e-01
-1.35884345e+00 2.10365295e-01 2.15272650e-01 -1.26141298e+00
2.30108704e-02 -1.38589203e-01 -6.20829284e-01 9.74765956e-01
-9.73147154e-01 -7.66055048e-01 3.79633784e-05 9.23449039e-01
3.81220043e-01 1.50413483e-01 1.12589395e+00 5.23607850e-01
-7.31747329e-01 4.78090614e-01 8.77024680e-02 -2.12526530e-01
7.43857980e-01 -1.02644229e+00 2.04250872e-01 4.68307585e-01
-5.26709497e-01 8.97719622e-01 6.27068520e-01 -1.04757655e+00
-1.30015242e+00 -1.10634756e+00 9.80947793e-01 -5.08196875e-02
7.40263760e-01 -2.86237508e-01 -9.07654703e-01 7.67841876e-01
-3.86063814e-01 2.03862578e-01 1.06045341e+00 6.84362650e-01
3.83754075e-02 4.88460250e-02 -1.08520651e+00 6.22057319e-01
1.23955989e+00 5.81394881e-03 1.45208582e-01 6.59405649e-01
2.70043224e-01 -3.29917610e-01 -1.17195559e+00 6.09243929e-01
5.63717008e-01 -6.30796969e-01 1.27429068e+00 -9.67324018e-01
-1.59716412e-01 -2.77533501e-01 3.58801693e-01 -1.01142776e+00
-8.38823199e-01 -8.84685099e-01 -1.64502040e-01 6.85260475e-01
6.15941763e-01 -7.77658880e-01 7.88269699e-01 8.26751649e-01
-1.00743093e-01 -7.37055182e-01 -1.34341609e+00 -5.41201174e-01
9.84621644e-02 -3.21840584e-01 3.38793546e-01 1.19436383e+00
1.45493820e-01 1.79121152e-01 -8.32076430e-01 2.03567743e-01
8.68736088e-01 -7.39463568e-02 4.67463821e-01 -1.36940527e+00
-3.63699347e-01 -1.39059260e-01 -2.23761410e-01 -6.78867579e-01
-2.42210738e-02 -7.83977151e-01 -4.19574350e-01 -1.36000180e+00
5.20197272e-01 -1.29911017e+00 -3.55424970e-01 7.26492822e-01
-3.78706872e-01 -2.33596399e-01 -5.10990500e-01 2.68270642e-01
-2.87734997e-02 3.76401454e-01 1.05758929e+00 2.55292244e-02
-6.18443072e-01 5.72221994e-01 -5.93463182e-01 5.49932241e-01
8.41846824e-01 -9.88401055e-01 -4.75630313e-01 -9.43520516e-02
1.91290285e-02 8.64788353e-01 3.74414831e-01 -4.63825613e-01
9.33787897e-02 -5.14139891e-01 4.02496189e-01 -5.45593143e-01
1.76021025e-01 -7.83073545e-01 8.55478525e-01 8.50954175e-01
-4.00696576e-01 3.60109121e-01 1.51832685e-01 6.83580577e-01
-3.61405537e-02 8.91236514e-02 6.33748412e-01 -8.15218687e-02
-1.52655065e-01 7.19569147e-01 -8.05988669e-01 3.26908641e-02
8.93415511e-01 -1.03313811e-01 -1.27036542e-01 -4.15576696e-01
-1.39620078e+00 6.93192005e-01 2.98980922e-01 -3.07212304e-02
7.02331364e-01 -1.31207156e+00 -9.32387233e-01 2.92198360e-01
7.14861751e-02 -1.51622459e-01 2.70213276e-01 1.35357988e+00
-3.37827891e-01 1.10812649e-01 1.54368296e-01 -6.51277006e-01
-1.65265250e+00 7.01382399e-01 4.23415065e-01 -4.72030967e-01
-8.54679048e-01 5.19784391e-01 4.40541327e-01 -3.71197686e-02
1.79132089e-01 -1.79437011e-01 -5.29864095e-02 8.96095335e-02
4.01350230e-01 5.54601550e-01 -1.05379455e-01 -4.54907231e-02
-6.81660295e-01 4.78273839e-01 2.36557171e-01 8.66494402e-02
1.28056037e+00 -1.96848810e-01 -2.88278580e-01 2.38658965e-01
1.00764084e+00 -1.16416931e-01 -9.08480883e-01 -1.18048020e-01
1.92737564e-01 -4.75810647e-01 9.34851244e-02 -4.48090911e-01
-9.80478704e-01 4.21193302e-01 4.64024782e-01 -1.01929531e-02
1.12750435e+00 -2.44387299e-01 2.48519003e-01 -8.27098042e-02
5.66879213e-01 -4.22259659e-01 -5.51274598e-01 4.56858575e-01
7.98352957e-01 -9.88860607e-01 4.00449783e-01 -7.69727409e-01
-4.71491277e-01 9.74059880e-01 2.30608415e-02 -1.12264575e-02
1.21122682e+00 4.09447402e-01 -1.00899870e-02 -3.41783255e-01
-1.05651093e+00 1.85386091e-01 3.68524119e-02 6.04365885e-01
6.55821681e-01 2.18828052e-01 -1.10679662e+00 6.17881298e-01
3.39800984e-01 8.29079077e-02 4.63323146e-01 1.03636193e+00
3.44040096e-02 -1.34420335e+00 -5.40381551e-01 9.15010810e-01
-6.55244052e-01 -3.56300205e-01 1.39546230e-01 7.61878371e-01
-4.09758747e-01 8.86610389e-01 -1.98012754e-01 2.93742537e-01
3.46200824e-01 -7.08584934e-02 2.70281415e-02 -8.35222721e-01
-6.94684029e-01 5.21724582e-01 5.93669415e-01 -5.47892094e-01
-1.84250772e-01 -1.02493846e+00 -1.11712086e+00 -2.67966211e-01
-2.27347389e-01 3.41973335e-01 1.66683257e-01 4.55832571e-01
5.83591938e-01 4.57782328e-01 7.35423684e-01 -4.27639425e-01
-4.20864284e-01 -5.82736552e-01 -9.95260954e-01 -9.42465384e-03
5.91882408e-01 -7.53431737e-01 -5.99513769e-01 4.31541614e-02] | [7.728705883026123, 5.284756183624268] |
f4cdf106-2f8f-44d2-b6da-789ddd995956 | assessing-yolact-for-real-time-and-robust | 2103.15997 | null | https://arxiv.org/abs/2103.15997v2 | https://arxiv.org/pdf/2103.15997v2.pdf | Assessing YOLACT++ for real time and robust instance segmentation of medical instruments in endoscopic procedures | Image-based tracking of laparoscopic instruments plays a fundamental role in computer and robotic-assisted surgeries by aiding surgeons and increasing patient safety. Computer vision contests, such as the Robust Medical Instrument Segmentation (ROBUST-MIS) Challenge, seek to encourage the development of robust models for such purposes, providing large, diverse, and annotated datasets. To date, most of the existing models for instance segmentation of medical instruments were based on two-stage detectors, which provide robust results but are nowhere near to the real-time (5 frames-per-second (fps)at most). However, in order for the method to be clinically applicable, real-time capability is utmost required along with high accuracy. In this paper, we propose the addition of attention mechanisms to the YOLACT architecture that allows real-time instance segmentation of instrument with improved accuracy on the ROBUST-MIS dataset. Our proposed approach achieves competitive performance compared to the winner ofthe 2019 ROBUST-MIS challenge in terms of robustness scores,obtaining 0.313 MI_DSC and 0.338 MI_NSD, while achieving real-time performance (37 fps) | ['Sharib Ali', 'Gilberto Ochoa-Ruiz', 'Leonardo Chang', 'Juan Carlos Angeles Ceron'] | 2021-03-30 | null | null | null | null | ['real-time-instance-segmentation'] | ['computer-vision'] | [-3.91002186e-03 2.62730330e-01 -3.72318923e-01 1.57698598e-02
-8.46272647e-01 -6.63760781e-01 2.80982345e-01 1.79310888e-01
-8.14508736e-01 2.87108302e-01 -1.90442309e-01 -4.77112204e-01
7.15194270e-02 -1.16505161e-01 -7.11467803e-01 -4.87972528e-01
-1.09460065e-03 3.57833415e-01 3.07164848e-01 -9.37595889e-02
2.81182081e-01 6.81257725e-01 -1.23661041e+00 9.30026844e-02
8.42015803e-01 1.08013940e+00 2.23434389e-01 9.67811465e-01
8.02651793e-02 4.32447970e-01 -1.64407760e-01 -4.35611337e-01
5.88336885e-01 -2.14066282e-01 -7.27475524e-01 -3.48342359e-01
3.43550354e-01 -2.01436535e-01 -2.18060181e-01 1.02259994e+00
5.78221619e-01 -6.25882447e-02 3.69123518e-01 -7.79260457e-01
-2.66657546e-02 2.22032681e-01 -5.36423385e-01 2.79312611e-01
2.75615185e-01 3.67876530e-01 4.17895824e-01 -6.86755002e-01
9.96852279e-01 9.21981156e-01 6.65703237e-01 7.06651390e-01
-8.52946699e-01 -5.85052073e-01 -1.26463383e-01 -3.48394096e-01
-9.11303043e-01 -3.50349039e-01 3.96760225e-01 -4.36132938e-01
5.13360798e-01 3.77037078e-01 7.28679538e-01 7.62178004e-01
7.70138204e-01 9.25614536e-01 7.76506782e-01 -2.07719341e-01
4.41391766e-02 4.14566696e-02 -1.92582488e-01 1.07443810e+00
5.30962706e-01 2.36189827e-01 -3.14825714e-01 2.84452975e-01
1.27063465e+00 3.80177535e-02 -3.35512727e-01 -6.76827967e-01
-1.84222329e+00 5.39067507e-01 6.75584078e-01 5.45603573e-01
-5.23170292e-01 2.74592042e-01 7.13547111e-01 -2.39479199e-01
7.57432834e-04 9.03363824e-01 -2.97674805e-01 -4.92058039e-01
-8.17324162e-01 -1.15151793e-01 7.46447921e-01 1.12672472e+00
9.08125285e-03 -1.45777509e-01 -3.12567472e-01 2.29303494e-01
9.60457250e-02 6.40708879e-02 5.97853541e-01 -1.17758870e+00
3.03406745e-01 6.39860392e-01 2.86922336e-01 -8.82361710e-01
-9.69098151e-01 -6.65540040e-01 -8.15615475e-01 2.67033428e-01
5.17982066e-01 -9.31449011e-02 -1.12681925e+00 1.17616296e+00
3.02362949e-01 2.59760827e-01 -1.19352914e-01 1.20171010e+00
9.97955203e-01 2.10186526e-01 1.89840451e-01 -1.83513954e-01
1.31596589e+00 -1.08731997e+00 -5.92844248e-01 -2.24158511e-01
7.93210864e-01 -9.32837129e-01 9.42210555e-01 5.28324544e-01
-1.25458717e+00 -5.49617648e-01 -1.00889504e+00 -7.19843879e-02
1.47484839e-01 3.41733396e-01 8.33562076e-01 7.25648046e-01
-7.86922514e-01 5.64349234e-01 -1.30660713e+00 -4.26053941e-01
5.52760363e-01 7.45486498e-01 -4.62505817e-01 5.72990067e-02
-3.67591530e-01 9.69706774e-01 3.18840623e-01 2.60983557e-01
-7.27269709e-01 -8.20621312e-01 -9.13676262e-01 -2.81766713e-01
3.55655581e-01 -6.75435960e-01 1.37429035e+00 -7.51808286e-01
-1.58211219e+00 1.16702569e+00 2.33015105e-01 -6.09122276e-01
9.12840068e-01 -4.31235880e-01 -3.78167927e-02 3.29251885e-01
-2.96877831e-01 8.79839242e-01 5.35015702e-01 -1.14759755e+00
-7.75326669e-01 -3.76694858e-01 3.54759872e-01 1.98159203e-01
-1.06544510e-01 -6.45084456e-02 -1.04234886e+00 -4.84998614e-01
2.73582190e-01 -1.44636548e+00 -6.46302879e-01 5.17604291e-01
-2.55666703e-01 2.58134365e-01 4.79751229e-01 -7.17496097e-01
9.57344592e-01 -2.21816087e+00 -3.92714562e-03 -2.52710670e-01
1.88418344e-01 6.10987544e-01 6.03083335e-03 -2.94903517e-01
3.18087548e-01 -1.87093571e-01 4.76893038e-02 -2.97611564e-01
-5.12901545e-01 2.39407897e-01 1.51694655e-01 8.06746781e-01
-2.13058367e-01 1.16448283e+00 -1.03447580e+00 -7.88122594e-01
7.30339050e-01 3.62746060e-01 -6.48906052e-01 2.75050551e-01
1.57201394e-01 9.65014696e-01 -4.48944688e-01 8.45060587e-01
4.93437529e-01 -2.34932855e-01 1.04503646e-01 -4.36172009e-01
-1.90257177e-01 -2.44768471e-01 -9.27142441e-01 2.37545681e+00
-6.84522808e-01 4.58430231e-01 3.45882028e-01 -5.70634961e-01
6.49844646e-01 4.32566762e-01 9.50951815e-01 -6.96008444e-01
4.53434378e-01 3.98185015e-01 1.78645149e-01 -6.17793143e-01
5.48736215e-01 5.04383035e-02 -1.21591669e-02 -2.90823966e-01
6.92323446e-02 -3.35728019e-01 1.83930069e-01 1.21300168e-01
7.52284348e-01 4.61062104e-01 3.36785555e-01 -4.82817322e-01
4.08640683e-01 3.45255136e-01 5.54783225e-01 6.66904390e-01
-6.69825792e-01 9.09358084e-01 2.21217304e-01 -5.04373729e-01
-8.10168445e-01 -7.10172176e-01 -1.01900220e-01 5.89197397e-01
8.20229292e-01 7.80709386e-02 -6.54071927e-01 -7.98023641e-01
-1.66277602e-01 2.34810814e-01 -5.62484086e-01 -8.04693997e-02
-6.74267292e-01 -3.10872048e-01 4.28943723e-01 5.69623291e-01
1.29102930e-01 -1.03681338e+00 -1.34715700e+00 4.45461154e-01
-1.55324027e-01 -1.35733485e+00 -4.74667341e-01 1.91645503e-01
-1.10242391e+00 -1.32456493e+00 -1.02169955e+00 -7.37071574e-01
7.21831381e-01 1.40058219e-01 9.61819589e-01 6.72418773e-02
-8.75637650e-01 2.63126016e-01 -2.08603695e-01 -5.94802380e-01
-3.01664054e-01 1.54192239e-01 -1.53639957e-01 -4.74986374e-01
-3.61513376e-01 3.59800249e-01 -1.08348072e+00 3.32834780e-01
-8.66051316e-01 2.50596970e-01 8.86679530e-01 6.92405820e-01
5.26719689e-01 -7.09139764e-01 1.37655452e-01 -9.44868684e-01
6.34391457e-02 5.24289766e-03 -7.66134441e-01 1.06388412e-01
-6.37224555e-01 -2.96398729e-01 3.08742464e-01 -5.16464829e-01
-9.47118402e-01 3.19811046e-01 -1.69786349e-01 -6.17828608e-01
-6.75442815e-02 3.56696010e-01 6.28946126e-01 -6.52807534e-01
6.62096381e-01 -2.33194187e-01 2.27677926e-01 -1.94652095e-01
2.15484738e-01 3.35781574e-01 9.14066255e-01 -3.47281337e-01
3.45086545e-01 6.07487321e-01 1.25384554e-01 -4.14824247e-01
-7.63608515e-01 -1.02990878e+00 -5.66703320e-01 -2.29799524e-01
8.24654818e-01 -8.77046525e-01 -8.74895334e-01 9.86049697e-02
-8.96900058e-01 -1.57877535e-01 -1.92599058e-01 9.15200949e-01
-7.05648243e-01 1.94190577e-01 -8.87978315e-01 -6.98272645e-01
-6.28225207e-01 -1.53655469e+00 1.21672595e+00 5.39023101e-01
-2.01128572e-01 -8.56742740e-01 -2.36159071e-01 3.83310378e-01
4.94895458e-01 6.34109497e-01 3.34600210e-01 -3.55477393e-01
-5.35474956e-01 -5.41935086e-01 -2.38034636e-01 3.44754681e-02
1.72117949e-01 3.64819355e-02 -7.40804851e-01 -4.57869589e-01
-1.19574025e-01 -2.53796577e-02 5.70443809e-01 8.81082714e-01
1.25006092e+00 1.11755989e-01 -6.33370757e-01 9.80514288e-01
1.48053133e+00 2.68883616e-01 3.89330298e-01 4.25653428e-01
5.84939241e-01 2.94026166e-01 1.11906826e+00 2.91599274e-01
2.86797583e-01 5.63943267e-01 8.72169256e-01 -6.13444090e-01
-9.30698514e-02 2.86312044e-01 -2.37789422e-01 5.14179826e-01
-4.08022851e-01 1.91592708e-01 -9.64765608e-01 5.99199057e-01
-1.87456584e+00 -4.97851431e-01 -2.58033335e-01 2.45975423e+00
6.17624402e-01 2.19988242e-01 -3.51447389e-02 -2.35942692e-01
4.30941612e-01 -2.85065502e-01 -4.83859807e-01 -2.66693830e-01
3.21035475e-01 -5.55469505e-02 8.37200761e-01 2.26467595e-01
-1.30726755e+00 7.64124036e-01 6.02768135e+00 4.35720205e-01
-1.43702960e+00 -1.16635419e-01 6.59628093e-01 -4.35947292e-02
3.10946077e-01 -3.40076149e-01 -5.62943578e-01 3.28895330e-01
5.96365809e-01 -1.27159413e-02 8.27406198e-02 9.82307673e-01
1.56313807e-01 -3.30800772e-01 -1.03254044e+00 1.19154251e+00
-4.66793142e-02 -1.57068682e+00 -3.81012619e-01 4.05921228e-02
7.27891505e-01 -1.07562400e-01 -1.85429417e-02 3.60160172e-01
-1.52885243e-01 -9.49389815e-01 5.82773805e-01 5.41634083e-01
8.58634412e-01 -6.05760992e-01 1.10271060e+00 3.53857487e-01
-1.11641467e+00 5.04364865e-03 -1.05218500e-01 5.00404298e-01
6.41656518e-02 3.57723743e-01 -1.09658170e+00 7.38550127e-01
7.21525371e-01 6.10923171e-01 -4.40347433e-01 1.66760302e+00
2.39016756e-01 1.34720862e-01 -1.61428735e-01 1.85974896e-01
3.28793436e-01 6.02882914e-02 3.95425260e-01 1.40985537e+00
3.05407912e-01 9.66765061e-02 2.77533680e-01 2.40133435e-01
-1.29908711e-01 -5.26365228e-02 -3.06417763e-01 3.25027816e-02
5.23075499e-02 1.66467631e+00 -1.07261431e+00 -1.65310547e-01
-3.23516786e-01 9.44035828e-01 -1.62027881e-01 -5.64840399e-02
-9.55912948e-01 -7.50891119e-02 3.84170175e-01 4.96036233e-03
-1.70420725e-02 -4.65011120e-01 -6.33986771e-01 -8.58983517e-01
-2.17886597e-01 -6.56371891e-01 3.96609157e-01 -5.32492459e-01
-6.47472024e-01 4.36203390e-01 -6.18621349e-01 -1.64677727e+00
-2.51407921e-01 -8.48385870e-01 -1.61363646e-01 4.66145664e-01
-1.71784759e+00 -1.38678241e+00 -6.18273675e-01 4.88170713e-01
6.95641816e-01 3.53621095e-01 8.39264333e-01 3.12938154e-01
-3.70433539e-01 5.41591227e-01 -5.14587760e-02 1.08340532e-01
8.68646502e-01 -1.17694056e+00 1.10042453e-01 6.97434962e-01
-8.11631829e-02 6.80240571e-01 7.34996557e-01 -3.51956487e-01
-1.91651654e+00 -1.07345021e+00 2.41285235e-01 -5.32181501e-01
2.54247695e-01 2.64223814e-02 -5.45637369e-01 5.94337463e-01
-1.24536268e-01 4.92938191e-01 3.51976901e-01 -2.18802840e-01
3.18287164e-01 1.57996193e-01 -1.30418932e+00 5.52758515e-01
8.68160963e-01 1.37242258e-01 -3.04580361e-01 2.64278859e-01
6.00546300e-01 -1.48099017e+00 -9.34203148e-01 7.23123610e-01
7.77192235e-01 -1.06552184e+00 1.04870176e+00 -2.10912317e-01
3.36369187e-01 -2.45202303e-01 5.48036873e-01 -7.02334881e-01
-1.39502138e-02 -6.71543777e-01 -1.91020906e-01 4.47611630e-01
2.01637805e-01 -2.16302231e-01 9.89063323e-01 6.58514857e-01
-3.97594064e-01 -8.31559181e-01 -1.09151232e+00 -7.18212843e-01
-3.40900749e-01 -4.27371919e-01 3.98542993e-02 8.09406877e-01
-7.83094242e-02 -4.66271281e-01 -8.72290209e-02 1.86035827e-01
4.76301819e-01 2.62098432e-01 8.37339640e-01 -1.13212872e+00
9.91727412e-02 -6.10464334e-01 -6.47889316e-01 -7.79380798e-01
-4.66130346e-01 -6.85874820e-01 2.13563487e-01 -1.80793917e+00
3.73319983e-02 -5.13488650e-01 -4.62237775e-01 2.81013429e-01
-3.55989903e-01 5.03191829e-01 3.54293197e-01 2.69006670e-01
-7.78370500e-01 2.68981252e-02 1.55578434e+00 7.24913105e-02
-3.39580268e-01 2.23609686e-01 -4.64723617e-01 7.90526986e-01
4.14672673e-01 -2.50252783e-01 1.85277313e-01 -3.40884298e-01
-2.09134221e-01 3.46503496e-01 2.57746935e-01 -1.23948479e+00
3.97337049e-01 -1.21432608e-02 4.38228279e-01 -3.08193773e-01
3.60207140e-01 -9.81590927e-01 -5.25977612e-02 1.14800525e+00
-1.48198369e-03 -2.51404971e-01 3.97521764e-01 3.55925262e-01
-2.17334285e-01 -6.57290220e-04 1.00166214e+00 -2.94343323e-01
-8.82587254e-01 4.35487568e-01 9.31238681e-02 -1.41186148e-01
1.33912766e+00 -4.81835961e-01 3.89736816e-02 -1.27882753e-02
-6.96785450e-01 2.17890248e-01 5.24563074e-01 7.23694861e-01
5.21256626e-01 -7.73929536e-01 -4.97753412e-01 -2.41627824e-02
3.80652040e-01 3.63925159e-01 3.80724847e-01 1.34994471e+00
-1.05183125e+00 7.55854547e-01 -1.66840985e-01 -8.96669447e-01
-1.10532451e+00 6.18709683e-01 4.80764478e-01 -3.52719486e-01
-7.90546894e-01 1.03710043e+00 9.97975990e-02 -3.26713204e-01
3.28286767e-01 -5.68624914e-01 -1.09626815e-01 -2.79360741e-01
2.35998094e-01 2.44353339e-01 3.28347117e-01 -2.95298070e-01
-4.66572076e-01 6.48870289e-01 -5.59146469e-03 1.99756563e-01
1.02533567e+00 3.17083020e-03 3.55589539e-01 1.39251649e-01
7.91473150e-01 8.28439370e-03 -1.41372359e+00 5.46459220e-02
9.74034369e-02 -5.82978189e-01 1.47110030e-01 -8.99627805e-01
-1.11047947e+00 6.60821199e-01 1.05417752e+00 -3.11962515e-01
9.13846195e-01 -1.73381239e-01 7.59789169e-01 -2.48427466e-02
7.88485050e-01 -7.99805224e-01 -1.28178805e-01 2.05171287e-01
6.44174397e-01 -1.54271662e+00 -3.94159518e-02 -4.72092807e-01
-7.02352107e-01 1.21640897e+00 7.20087171e-01 -9.90765318e-02
1.61137894e-01 4.56239372e-01 4.33828235e-01 1.14669492e-02
-7.32877776e-02 -5.49381413e-02 6.41888142e-01 3.04650217e-01
6.58805072e-01 1.77220590e-02 -4.80842263e-01 2.96100229e-01
1.04774870e-01 3.07493180e-01 4.13804114e-01 1.11661053e+00
-2.46206477e-01 -6.07372105e-01 -2.10349634e-01 5.02472878e-01
-7.74921894e-01 2.76705831e-01 4.44979757e-01 9.57110703e-01
-4.46922407e-02 8.27046156e-01 -1.23931661e-01 2.13613942e-01
5.30405223e-01 -4.82933551e-01 5.29940963e-01 -5.84293723e-01
-9.58598375e-01 1.99423417e-01 -1.12281732e-01 -1.05707920e+00
-3.36238533e-01 -3.17218125e-01 -1.56383836e+00 2.76452713e-02
-3.38852853e-01 -7.74559975e-02 1.20067847e+00 6.62871122e-01
3.49482119e-01 8.52110326e-01 3.47698182e-01 -1.26459646e+00
-2.98370719e-01 -6.08413219e-01 -3.12554687e-01 3.89167488e-01
4.32296395e-01 -3.29008669e-01 8.45501348e-02 1.40117541e-01] | [14.049773216247559, -3.207242012023926] |
8277b287-b5e0-4455-aba4-518e07d5272d | document-ai-benchmarks-models-and | 2111.08609 | null | https://arxiv.org/abs/2111.08609v1 | https://arxiv.org/pdf/2111.08609v1.pdf | Document AI: Benchmarks, Models and Applications | Document AI, or Document Intelligence, is a relatively new research topic that refers to the techniques for automatically reading, understanding, and analyzing business documents. It is an important research direction for natural language processing and computer vision. In recent years, the popularity of deep learning technology has greatly advanced the development of Document AI, such as document layout analysis, visual information extraction, document visual question answering, document image classification, etc. This paper briefly reviews some of the representative models, tasks, and benchmark datasets. Furthermore, we also introduce early-stage heuristic rule-based document analysis, statistical machine learning algorithms, and deep learning approaches especially pre-training methods. Finally, we look into future directions for Document AI research. | ['Furu Wei', 'Tengchao Lv', 'Yiheng Xu', 'Lei Cui'] | 2021-11-16 | null | null | null | null | ['document-image-classification', 'document-layout-analysis', 'document-ai'] | ['computer-vision', 'computer-vision', 'natural-language-processing'] | [ 3.42832297e-01 -2.98698813e-01 -4.42835093e-01 -2.53064930e-01
-2.62477309e-01 -5.97998500e-01 1.06240392e+00 5.86855590e-01
1.17132358e-01 3.56157988e-01 2.04728693e-01 -5.79666376e-01
-2.48169765e-01 -8.68347764e-01 -2.33855247e-01 -5.99613130e-01
9.28232670e-02 5.41138947e-01 -1.45228624e-01 1.71675906e-01
1.02895379e+00 6.57635152e-01 -1.27279866e+00 8.66446853e-01
6.26118660e-01 1.14984369e+00 1.84722953e-02 1.12116587e+00
-9.13700998e-01 1.34170556e+00 -9.22748148e-01 -3.26768577e-01
-5.23919463e-01 -2.46132836e-01 -9.84108984e-01 5.37890673e-01
4.00178403e-01 -5.53598464e-01 -4.05587703e-01 9.93735135e-01
1.58470884e-01 3.12331636e-02 1.08034885e+00 -1.10969663e+00
-1.52865303e+00 6.82793021e-01 -8.05263102e-01 2.83281446e-01
2.81340420e-01 -1.71841785e-01 8.92003596e-01 -8.44126880e-01
7.49173820e-01 1.41224623e+00 1.65579945e-01 3.56147826e-01
-4.88392442e-01 1.38679538e-02 4.77482975e-01 7.45494843e-01
-1.02924955e+00 6.58972785e-02 1.04140723e+00 -6.20454550e-01
1.08720255e+00 6.11603931e-02 5.73385179e-01 1.03102434e+00
4.92571682e-01 1.91304755e+00 3.64212662e-01 -1.08663857e+00
1.87429622e-01 -7.17018470e-02 8.28011870e-01 9.52547073e-01
1.89372942e-01 -3.40774596e-01 -1.23531051e-01 2.58823782e-01
4.91335720e-01 1.91076189e-01 1.11836597e-01 -3.39172632e-01
-9.64388311e-01 1.00013423e+00 2.44306892e-01 7.95113325e-01
-2.32214063e-01 -7.11740255e-02 9.19754565e-01 -3.68699171e-02
7.05713704e-02 4.77619827e-01 -1.12705708e-01 -6.28265971e-03
-8.60534787e-01 3.57143968e-01 5.29805958e-01 1.05494833e+00
1.75233688e-02 5.13324738e-01 -6.46960497e-01 9.04316962e-01
3.24911267e-01 3.46349120e-01 6.46361470e-01 -4.93078589e-01
4.86610800e-01 9.60391164e-01 -2.29196236e-01 -1.24703765e+00
-3.95825893e-01 6.31990880e-02 -1.12923229e+00 -7.02500716e-02
-1.94358509e-02 1.15332276e-01 -1.36908412e+00 4.85764503e-01
-3.90999168e-01 -9.02696967e-01 2.13414412e-02 5.75846553e-01
1.05406880e+00 1.22841513e+00 -6.13424070e-02 -2.07677454e-01
1.35033524e+00 -1.18714142e+00 -1.05177009e+00 -2.67260402e-01
4.10512596e-01 -7.06357598e-01 1.02410007e+00 9.95386958e-01
-9.81104374e-01 -5.58093190e-01 -1.07056773e+00 -5.45930326e-01
-1.09246457e+00 5.37113905e-01 1.01078248e+00 3.71868998e-01
-5.88539362e-01 -3.34438235e-02 -5.55967748e-01 -5.92548490e-01
9.33808029e-01 1.56279847e-01 1.43397599e-01 -4.94228750e-01
-7.89630890e-01 6.93525851e-01 7.41478682e-01 1.03492506e-01
-7.27143407e-01 -1.82010546e-01 -7.34853804e-01 3.79121721e-01
4.71565127e-01 -1.98923603e-01 1.37464392e+00 -8.56340766e-01
-1.24108505e+00 8.69754255e-01 -1.54134408e-01 -5.41018963e-01
4.06753160e-02 -2.53075182e-01 -7.41437852e-01 9.49237719e-02
-2.77461380e-01 4.60135072e-01 9.27821994e-01 -1.37054133e+00
-6.28878236e-01 -6.83513641e-01 -1.60904542e-01 -4.27135639e-02
-5.05172312e-01 3.15113246e-01 -7.05472171e-01 -5.66459596e-01
-2.17057526e-01 -3.20106119e-01 2.10429411e-02 1.00519821e-01
-7.55691290e-01 -6.94340408e-01 1.41799080e+00 -6.65776074e-01
1.18400645e+00 -2.18708944e+00 1.15804181e-01 4.33025181e-01
8.15364122e-02 5.56073010e-01 -1.21845648e-01 4.90028054e-01
2.15476185e-01 2.44830847e-01 -1.40500933e-01 3.45932424e-01
3.77497748e-02 2.10072026e-02 -6.91781640e-01 -9.51234177e-02
6.88181296e-02 1.26337123e+00 -6.20556235e-01 -7.66616642e-01
4.72036302e-01 2.26841316e-01 1.88769251e-01 -9.42663327e-02
-7.72979617e-01 -4.48000818e-01 -6.44249141e-01 1.04714835e+00
4.01468456e-01 -3.24023068e-01 2.16872230e-01 -5.95136508e-02
-6.46314397e-02 -3.12526934e-02 -8.53587747e-01 1.46714222e+00
-1.06538318e-01 1.53820944e+00 -2.95675009e-01 -1.47833037e+00
1.10151887e+00 4.63130185e-03 2.69302994e-01 -1.08669531e+00
3.83999437e-01 -1.66576043e-01 -2.87290961e-02 -8.14793646e-01
7.59559929e-01 5.63540459e-01 2.61577308e-01 6.79185271e-01
-3.31200808e-01 -2.23837793e-01 8.81414235e-01 2.26761207e-01
6.48921847e-01 -2.13531598e-01 5.85509717e-01 1.33410364e-03
8.68702412e-01 4.02776539e-01 -3.61260056e-01 6.74999535e-01
7.65044838e-02 3.33329320e-01 4.55197155e-01 -9.34427738e-01
-9.67614889e-01 -8.31564546e-01 -2.00634860e-02 1.21546483e+00
-4.57030088e-02 -5.07269800e-01 -8.13310504e-01 -6.55289888e-01
2.10865960e-01 7.23329008e-01 -5.76610088e-01 -1.23398043e-01
-7.00959623e-01 -6.07610822e-01 5.52160740e-01 1.15506852e+00
7.74058878e-01 -1.65562451e+00 -2.35266730e-01 3.89952846e-02
3.02994639e-01 -7.03776360e-01 -8.75648670e-03 1.03186242e-01
-8.71210933e-01 -1.03757477e+00 -7.35086501e-01 -1.25303519e+00
7.70152569e-01 3.91377419e-01 1.00310445e+00 2.85680562e-01
-6.36766434e-01 3.22358459e-01 -4.76524949e-01 -9.04106081e-01
-3.24816465e-01 6.46772981e-02 -4.50440407e-01 -4.83692735e-01
9.19426024e-01 3.70117247e-01 -2.34619081e-01 -3.36745948e-01
-1.16386080e+00 -1.37973085e-01 6.98905289e-01 7.68807054e-01
5.42939544e-01 5.38320601e-01 3.83473113e-02 -9.77500856e-01
1.40390396e+00 7.50239864e-02 -4.28467691e-01 8.45583856e-01
-8.02641571e-01 8.43835473e-02 8.16451609e-01 -2.50104934e-01
-1.20112467e+00 -5.25778413e-01 4.22185510e-01 1.65935919e-01
-4.74942774e-01 8.52531016e-01 -3.63850355e-01 3.85226697e-01
6.45820081e-01 6.13893747e-01 -3.22208434e-01 -3.75888646e-01
5.73256373e-01 1.13453794e+00 4.83360767e-01 -4.70218301e-01
4.20581102e-01 4.84257936e-01 -2.00009197e-01 -9.54431057e-01
-6.88768506e-01 -1.38349071e-01 -8.81227493e-01 -1.41702056e-01
9.74767745e-01 -2.99142692e-02 -7.71667063e-01 4.86659437e-01
-1.48757589e+00 -5.24393879e-02 -8.56258199e-02 2.03251652e-02
-2.24197626e-01 4.94953185e-01 -7.26037323e-01 -7.06703305e-01
-7.07724750e-01 -9.89238203e-01 1.05069613e+00 4.04386431e-01
-3.40133131e-01 -1.15081775e+00 -6.81342781e-02 3.89974356e-01
1.67153612e-01 2.41952408e-02 1.67650771e+00 -5.81868291e-01
-4.66165990e-01 -5.33681333e-01 -6.71673119e-01 3.62445921e-01
1.20870389e-01 7.99530685e-01 -6.81208491e-01 -5.45162708e-02
-4.93431240e-01 -4.43770558e-01 1.12006485e+00 6.67461276e-01
2.04656601e+00 -2.90112257e-01 -6.61888003e-01 1.97728485e-01
1.16282582e+00 1.16017103e+00 5.88598847e-01 6.27733827e-01
9.71819878e-01 5.21175027e-01 4.63760793e-01 3.60768676e-01
1.71361476e-01 1.23009183e-01 8.08807984e-02 6.67255372e-02
1.51312441e-01 -2.54325513e-02 1.02345943e-01 6.46181107e-01
-1.51877478e-01 -9.73357201e-01 -1.27240300e+00 3.22752565e-01
-1.87004423e+00 -1.04827964e+00 -2.32816748e-02 1.34613502e+00
2.61821657e-01 3.29336017e-01 -2.28148967e-01 7.59432912e-01
6.47116840e-01 3.30933303e-01 -6.68340504e-01 -1.08653116e+00
-2.06413239e-01 3.96166220e-02 -2.80897096e-02 5.26105762e-02
-1.28776693e+00 1.15336287e+00 6.81616020e+00 6.87860847e-01
-1.05100918e+00 -6.98642969e-01 6.65782690e-01 3.27644974e-01
-2.03754991e-01 -4.46478099e-01 -5.80438495e-01 2.25548506e-01
4.62620229e-01 -2.58069694e-01 3.46550345e-01 1.25571144e+00
-1.35097755e-02 -1.47129297e-01 -1.21423900e+00 9.40303266e-01
6.91839874e-01 -1.96514726e+00 1.00750148e+00 1.17411371e-02
7.15205312e-01 -7.14990377e-01 1.92044899e-01 2.52116024e-01
2.55175292e-01 -1.21526742e+00 5.27405441e-01 4.48171884e-01
4.33373302e-01 -1.13278151e+00 6.72506452e-01 1.08051538e-01
-1.02840996e+00 -4.15470451e-01 -4.51385647e-01 1.09759644e-02
-1.07340105e-01 3.64791840e-01 -6.02680624e-01 3.72875333e-01
7.23594189e-01 1.09306586e+00 -7.82667696e-01 8.15194666e-01
-1.46344811e-01 4.38745737e-01 4.07610923e-01 -6.82513297e-01
6.17668867e-01 -1.27131775e-01 8.62208381e-02 1.42887068e+00
-2.00296268e-01 -8.26761574e-02 1.34778172e-01 7.65326679e-01
-1.52032092e-01 1.34138271e-01 -5.67969799e-01 -7.24715412e-01
4.93886732e-02 9.69601095e-01 -1.16782689e+00 -5.92185557e-01
-4.77828771e-01 8.13919663e-01 -3.16510838e-03 5.16494513e-01
-3.79047692e-01 -9.61151361e-01 -5.96048636e-03 -2.17114478e-01
1.96936324e-01 -4.80764866e-01 -6.88058972e-01 -8.02328587e-01
8.66391659e-02 -1.15208316e+00 5.68703532e-01 -1.01259077e+00
-1.00425160e+00 3.60022068e-01 -1.02895468e-01 -8.70907843e-01
-4.76034641e-01 -1.47621310e+00 -8.60450685e-01 1.58178747e-01
-1.13822198e+00 -1.07788336e+00 -3.84039700e-01 4.30204391e-01
1.15155900e+00 -7.73345292e-01 7.52813995e-01 -1.73365116e-01
-7.57104218e-01 2.80513257e-01 7.23673880e-01 5.67475975e-01
1.76436096e-01 -1.41549027e+00 4.20231164e-01 7.74020612e-01
5.65763593e-01 5.71120679e-01 1.99553862e-01 -3.39199960e-01
-1.52273583e+00 -7.41077304e-01 5.47036946e-01 -2.11287990e-01
4.71399486e-01 -4.29949731e-01 -9.28191602e-01 6.44225419e-01
5.71234584e-01 -4.94699210e-01 6.13564551e-01 6.29867241e-03
-1.43718779e-01 -1.90469787e-01 -8.55061889e-01 8.40520442e-01
5.08448422e-01 -4.69033182e-01 -5.50078332e-01 6.23935163e-01
4.09697473e-01 -5.13906330e-02 -3.14603031e-01 -1.74421936e-01
5.25469363e-01 -6.23559594e-01 8.00795794e-01 -1.01465368e+00
7.70623088e-01 -6.02262542e-02 5.37496107e-03 -7.98196554e-01
-5.36950469e-01 -1.80676207e-01 -7.57323921e-01 1.21977627e+00
9.61449072e-02 3.19273025e-01 9.52977121e-01 4.24135208e-01
-5.86852096e-02 -7.44447708e-01 -1.69131771e-01 -5.69661140e-01
1.99349493e-01 -5.17162263e-01 6.30191267e-01 7.65857637e-01
9.31798741e-02 4.54540014e-01 4.10344787e-02 -4.44352627e-01
2.65247613e-01 5.15415967e-01 5.72903216e-01 -1.45170438e+00
2.47474357e-01 -1.07664442e+00 -3.28399777e-01 -1.26496422e+00
2.31961474e-01 -7.94783473e-01 -1.70158923e-01 -2.43942761e+00
2.63397336e-01 5.30166328e-01 -1.58940867e-01 3.51408452e-01
2.85072505e-01 -3.46727222e-01 8.55587889e-03 8.05244148e-02
-6.31268442e-01 3.17080617e-01 1.46004093e+00 -1.07887316e+00
-1.98439136e-01 -2.06791922e-01 -6.87725425e-01 9.13198709e-01
7.96491385e-01 1.55424714e-01 -4.86778736e-01 -9.10526276e-01
1.73371404e-01 -5.23263991e-01 -2.72102486e-02 -5.74294508e-01
2.76345462e-01 -4.28361893e-01 1.21176100e+00 -1.14938080e+00
-1.14187919e-01 -7.32902706e-01 -1.10079813e+00 4.06748950e-01
-7.87182271e-01 7.76874721e-02 2.53625095e-01 3.39169264e-01
-5.00468314e-01 -5.24663866e-01 4.43734616e-01 -2.46142358e-01
-1.11409044e+00 1.29621625e-01 -1.10743618e+00 -1.87048823e-01
1.06188715e+00 -4.96781737e-01 -6.03231668e-01 -8.19405541e-02
-1.95134535e-01 3.31487685e-01 -2.06127062e-01 8.28716457e-01
1.06567538e+00 -1.13257718e+00 -4.53882217e-01 -1.22525473e-03
3.31651956e-01 1.97440788e-01 -1.68109134e-01 -2.06595026e-02
-8.32269311e-01 1.11737621e+00 -2.90077567e-01 -4.40374941e-01
-1.39880860e+00 1.00344086e+00 -1.24201849e-01 -2.84882277e-01
-5.48370898e-01 6.63391471e-01 -6.03310391e-02 2.57559001e-01
7.66479671e-01 -5.43821037e-01 -7.44546771e-01 -3.38226482e-02
9.41210747e-01 5.59664309e-01 -9.76542477e-03 -4.57057580e-02
-4.59449172e-01 5.80723941e-01 -9.09577489e-01 2.06941694e-01
1.01244390e+00 2.29853138e-01 -2.77295262e-01 3.69518310e-01
1.00667381e+00 -4.54675108e-01 -4.48448986e-01 -4.42187414e-02
6.65975690e-01 -2.06496552e-01 2.69766808e-01 -8.72827470e-01
-9.60300922e-01 1.37134564e+00 3.49954844e-01 5.64779699e-01
9.70460236e-01 -1.00327499e-01 5.74271441e-01 1.32717943e+00
-1.82280868e-01 -1.59287286e+00 4.49627221e-01 7.12743342e-01
1.14614165e+00 -1.36466730e+00 4.88200009e-01 1.21070191e-01
-6.39384329e-01 1.71089435e+00 7.89979637e-01 -2.27146931e-02
7.02415109e-01 3.90657604e-01 -2.41769906e-02 -5.41612089e-01
-4.91638809e-01 1.12739047e-02 5.38609982e-01 8.17323387e-01
8.45929861e-01 -1.66257858e-01 -7.35558867e-02 4.49769080e-01
-8.63466784e-02 9.81242806e-02 3.09573889e-01 1.27443433e+00
-6.82662547e-01 -9.12370563e-01 -3.99195701e-01 7.37187743e-01
-1.39074743e-01 -8.42681900e-02 -1.16436267e+00 8.05577636e-01
-1.60248369e-01 7.15174615e-01 5.09310424e-01 -1.18015530e-04
1.55232430e-01 1.71994209e-01 6.69247210e-01 -4.14267153e-01
-4.39452045e-02 -3.23075235e-01 -1.86761826e-01 -1.48576638e-02
-2.28483185e-01 -2.90017039e-01 -1.47323453e+00 -3.69938582e-01
-2.06644073e-01 5.81762455e-02 7.35084057e-01 1.19744003e+00
1.70730069e-01 7.07545459e-01 2.80054539e-01 -4.10303742e-01
6.37726346e-03 -9.47986007e-01 -4.52458322e-01 -8.57528076e-02
1.51672974e-01 -4.21872199e-01 2.52299815e-01 5.25707304e-01] | [11.617058753967285, 2.6606860160827637] |
d90b011d-0690-41b5-b96f-aadd10681d35 | lidar-based-3d-tracking-and-state-estimation | 2304.01396 | null | https://arxiv.org/abs/2304.01396v1 | https://arxiv.org/pdf/2304.01396v1.pdf | Lidar based 3D Tracking and State Estimation of Dynamic Objects | State estimation of oncoming vehicles: Earlier research has been based on determining states like position, velocity, orientation , angular velocity, etc of ego-vehicle. Our approach focuses on estimating the states of non-ego vehicles which is crucial for Motion planning and decision-making. Dynamic Scene Based Localization: Our project will work on dynamic scenes like moving ego (self) and non-ego vehicles. Previous methods were focused on static environments. | ['Gautham Narayan Narasimhan', 'Patil Shubham Suresh'] | 2023-04-03 | null | null | null | null | ['motion-planning'] | ['robots'] | [-5.83109140e-01 -2.68595874e-01 -3.15941840e-01 -4.55725014e-01
1.44288875e-02 -7.16432154e-01 9.14357245e-01 -1.26485690e-01
-3.18798244e-01 5.94688058e-01 3.15541834e-01 -3.41065586e-01
4.53304797e-01 -7.25698531e-01 -3.56592268e-01 -4.60627466e-01
-5.41029215e-01 4.90301400e-01 7.72912383e-01 -1.98095605e-01
5.34907520e-01 9.71548676e-01 -1.26544046e+00 -1.98134646e-01
5.69712408e-02 2.86278337e-01 1.89295679e-01 1.40682399e+00
-5.94510064e-02 9.22248781e-01 -1.89555347e-01 7.67791197e-02
1.45907000e-01 -1.52654946e-01 -5.57444274e-01 1.21330554e-02
2.34970629e-01 -5.28148293e-01 -8.13404799e-01 1.02076960e+00
2.24175930e-01 3.83647174e-01 4.81311232e-01 -1.60040581e+00
1.47644863e-01 6.64136857e-02 -2.67513007e-01 6.86500490e-01
5.76337814e-01 1.56728402e-01 1.39262095e-01 -4.98189837e-01
1.04496944e+00 1.10445738e+00 6.11084759e-01 4.77119625e-01
-7.32992887e-01 -5.93350053e-01 1.66550383e-01 9.91571069e-01
-1.56025469e+00 -1.08394051e+00 7.31202364e-01 -3.74726206e-01
1.27391064e+00 6.37121722e-02 6.53356969e-01 6.40409052e-01
7.79388070e-01 5.97932935e-01 8.31012905e-01 -1.21733278e-01
5.05875170e-01 3.01601529e-01 1.99879736e-01 5.15085161e-01
2.75715679e-01 1.54204428e-01 -2.33695030e-01 1.15265600e-01
6.58932328e-01 -2.01250777e-01 1.75554186e-01 -9.00623858e-01
-1.08116329e+00 7.40740001e-01 1.12320706e-01 3.74511987e-01
-8.26067746e-01 4.65579212e-01 3.64073008e-01 2.35676542e-01
1.36833906e-01 -2.07079545e-01 -1.84167922e-01 -5.74076235e-01
-8.69570971e-01 2.98158854e-01 7.32377410e-01 1.25449026e+00
1.17611015e+00 2.88746536e-01 1.14982359e-01 -9.85018462e-02
4.08888042e-01 6.94534719e-01 1.75015107e-01 -1.19869542e+00
3.02396744e-01 3.41283560e-01 4.85299885e-01 -1.15198600e+00
-6.51172280e-01 7.25453794e-02 -1.37427732e-01 2.62266845e-01
-1.06212705e-01 -6.86646402e-01 -1.07164168e+00 1.66238213e+00
6.30274713e-01 7.24020600e-01 3.55561614e-01 6.81439459e-01
8.27289581e-01 5.68944037e-01 2.56151795e-01 -2.77089536e-01
8.12343776e-01 -7.60516405e-01 -9.30141985e-01 -5.20916700e-01
6.64077938e-01 -7.33913660e-01 -3.71584624e-01 -2.88785875e-01
-9.11876142e-01 -5.03352106e-01 -8.47295046e-01 4.14589673e-01
-5.99415183e-01 -1.68102130e-01 6.66690946e-01 5.97651064e-01
-1.60906148e+00 2.29240075e-01 -1.10003531e+00 -1.02082491e+00
-3.78757864e-02 5.58074236e-01 -7.81264305e-01 8.44991859e-03
-1.01325464e+00 1.41336811e+00 4.67128932e-01 -2.24101357e-02
-1.23549545e+00 7.63895065e-02 -1.24370992e+00 -4.57830638e-01
6.62380606e-02 -6.95100367e-01 1.10652769e+00 -5.03243506e-01
-1.44595158e+00 5.23586452e-01 -9.63383436e-01 -4.47567105e-01
2.58940458e-01 1.48116201e-01 -8.45315933e-01 4.59521189e-02
3.08747679e-01 5.22977293e-01 3.20841402e-01 -1.24527466e+00
-8.15323412e-01 -6.39760494e-01 1.00885801e-01 7.22318828e-01
5.37351787e-01 1.82858899e-01 -4.84925091e-01 4.31058794e-01
6.19242251e-01 -1.25007582e+00 -5.33339500e-01 -6.61591470e-01
-3.59832793e-01 -1.02094643e-01 1.52534282e+00 -1.91350356e-01
8.81533563e-01 -1.91902912e+00 -3.14242244e-01 -1.39335236e-02
-7.60521814e-02 2.61418611e-01 4.61040139e-02 7.05973923e-01
-6.84315637e-02 -4.25243020e-01 8.11759412e-01 -1.31544694e-01
-1.61607951e-01 2.21289724e-01 -6.30607828e-02 1.05517244e+00
-3.49175006e-01 7.87698209e-01 -1.05287707e+00 -7.88121164e-01
8.15355659e-01 4.51340944e-01 -3.14030677e-01 -4.77858223e-02
5.15795052e-01 3.04433227e-01 -8.08466494e-01 6.62698865e-01
8.02854538e-01 4.02382016e-01 -3.00025269e-02 4.17730920e-02
-8.71972620e-01 2.48388290e-01 -1.28850186e+00 1.28221023e+00
-3.56399745e-01 1.26895344e+00 2.56733149e-01 -6.01612091e-01
5.17861724e-01 4.34612006e-01 7.92888105e-01 -3.27529579e-01
3.91463190e-01 -2.80101538e-01 -1.14174895e-01 -6.75440729e-01
9.62901533e-01 6.37684539e-02 -3.98100577e-02 1.47300577e-02
-7.98183605e-02 -4.25757915e-02 3.74968886e-01 3.99961829e-01
1.34003961e+00 1.18287675e-01 4.17315394e-01 -1.03111506e-01
4.98348147e-01 5.14059961e-01 4.62079704e-01 6.03086472e-01
-9.92657423e-01 4.90744002e-02 8.35786387e-02 -3.53381455e-01
-1.00000966e+00 -1.13531792e+00 -3.30381915e-02 6.77459896e-01
8.39408875e-01 -3.80882531e-01 -6.36181712e-01 -1.90257296e-01
-1.70931354e-01 9.19632375e-01 -5.68687677e-01 7.17309909e-03
-7.02326953e-01 -8.63786042e-02 6.96203634e-02 4.14261729e-01
5.39065540e-01 -7.68523812e-01 -7.84489274e-01 5.25226831e-01
-9.68436059e-03 -1.42635465e+00 -2.31145620e-01 -1.83590323e-01
-7.40484178e-01 -9.24409151e-01 -2.37347856e-01 -5.29675305e-01
6.21089876e-01 8.03547919e-01 7.19469786e-01 -5.52696824e-01
2.04909146e-01 7.88642347e-01 -7.85875395e-02 -7.13183105e-01
-2.79959977e-01 -3.45424384e-01 5.77103198e-01 -9.89924967e-02
7.11664140e-01 -4.38476384e-01 -6.71579480e-01 5.77874601e-01
-2.77089309e-02 8.56371596e-02 2.41248071e-01 -2.95494884e-01
3.22627038e-01 1.88439623e-01 -6.13842495e-02 -5.78861475e-01
1.33512065e-01 -7.18218863e-01 -5.78490734e-01 -1.21424489e-01
-1.76842466e-01 -4.78320807e-01 1.86957255e-01 -1.32424310e-01
-9.36344743e-01 3.94526452e-01 -1.73941553e-02 -4.82845306e-01
-5.24437070e-01 -1.58444256e-01 -5.10255946e-03 -3.05330187e-01
2.78719246e-01 3.05496901e-01 -4.58734214e-01 2.94076949e-01
2.43687242e-01 4.96725082e-01 5.40332973e-01 2.74629176e-01
7.23503053e-01 9.50192094e-01 3.05249333e-01 -1.10941875e+00
2.52125934e-02 -1.18829894e+00 -8.45106840e-01 -6.31079257e-01
9.79291916e-01 -1.15061831e+00 -6.79489315e-01 3.25308532e-01
-1.25907755e+00 -2.31268108e-01 1.22388534e-01 9.34328020e-01
-7.15749919e-01 3.93749326e-01 -3.15486908e-01 -1.06449652e+00
4.47665229e-02 -1.06343997e+00 9.90356624e-01 5.05713642e-01
-1.37429893e-01 -1.33522725e+00 6.39269769e-01 -2.33570620e-01
6.70737684e-01 2.01149046e-01 1.68386865e-02 -3.35276127e-01
-8.63221586e-01 -6.34538591e-01 -1.00627802e-01 -6.28926575e-01
9.62437317e-02 1.75543696e-01 -5.15360415e-01 -1.45000443e-01
-2.27777138e-01 5.22833645e-01 3.93234938e-01 8.14945757e-01
5.61874011e-04 -1.17605619e-01 -9.73846793e-01 5.48211813e-01
1.55995715e+00 5.54698229e-01 6.45435095e-01 3.75855416e-01
5.86991966e-01 5.03946543e-01 8.91280890e-01 2.49106973e-01
9.38473463e-01 8.32023799e-01 5.95384657e-01 3.24211001e-01
-2.08388735e-02 -1.31252140e-01 5.28187156e-01 3.86810094e-01
-2.77776480e-01 -4.13872212e-01 -9.34333444e-01 8.91681612e-01
-1.86328363e+00 -1.56756413e+00 -5.08486688e-01 1.71131015e+00
-2.34084830e-01 -5.55075053e-03 -6.25683069e-02 -5.08132219e-01
8.70706499e-01 2.55655259e-01 -6.53612196e-01 -6.82809949e-01
2.59941757e-01 -7.46313989e-01 1.00063586e+00 7.27220833e-01
-1.38539076e+00 1.48701441e+00 8.04841995e+00 1.14252500e-01
-1.30651700e+00 1.53088942e-01 -8.88954010e-03 6.92472085e-02
-7.58108869e-02 4.48315889e-01 -1.29402423e+00 2.59751827e-01
1.11285508e+00 -6.70617402e-01 2.62987137e-01 1.24111021e+00
6.92608297e-01 -7.53066778e-01 -7.32621431e-01 8.12203646e-01
-2.82742158e-02 -1.16628063e+00 -4.36805427e-01 2.33132064e-01
9.03017044e-01 6.67986810e-01 -4.53892261e-01 3.89458209e-01
6.38517022e-01 -4.50257897e-01 5.57875335e-01 4.42266971e-01
4.13019001e-01 -6.35750234e-01 9.75294173e-01 4.55443978e-01
-1.41091275e+00 1.07414037e-01 -4.73137945e-01 -2.02774912e-01
8.52517188e-01 4.51303609e-02 -1.25777578e+00 2.26822719e-01
4.20508653e-01 5.51563621e-01 -3.20608497e-01 1.02794683e+00
8.03494453e-02 3.87231916e-01 -4.33199495e-01 -1.33211911e-01
5.32189727e-01 -3.42574567e-01 1.12870550e+00 1.50817728e+00
1.64954409e-01 3.58512044e-01 3.39852631e-01 2.88224727e-01
7.08801091e-01 -9.89813581e-02 -1.33640087e+00 2.08242998e-01
2.33270258e-01 1.31038511e+00 -8.79590631e-01 -6.34744167e-01
-1.91602036e-01 6.95030272e-01 -9.19603407e-02 3.66407603e-01
-9.46918070e-01 -2.00931966e-01 1.02248836e+00 3.73805255e-01
2.64047831e-01 -9.22705591e-01 1.72084779e-01 -1.04323781e+00
-6.11667097e-01 2.22990334e-01 -7.79305249e-02 -8.41404498e-01
-1.21847436e-01 6.03287935e-01 3.86762410e-01 -1.19893944e+00
-8.43061805e-01 -2.98196614e-01 -1.11160100e+00 6.58969104e-01
-1.34622467e+00 -1.12660301e+00 -3.63581777e-01 7.65370369e-01
7.10581064e-01 -3.11222583e-01 4.40788716e-01 1.59036949e-01
-5.15337527e-01 -1.50196299e-01 -1.34478167e-01 8.61580893e-02
2.67810464e-01 -8.84498835e-01 5.83359361e-01 1.25284803e+00
-1.41233400e-01 7.56796896e-01 1.44498181e+00 -8.98804724e-01
-1.87357092e+00 -9.97760773e-01 9.92553711e-01 -7.18109787e-01
7.30073392e-01 -3.15091908e-02 -2.09536076e-01 1.08308625e+00
1.26801014e-01 1.80238768e-01 1.53371826e-01 -1.18811823e-01
5.51754892e-01 2.06083730e-01 -9.70423102e-01 7.51508176e-01
9.10328150e-01 -5.45835376e-01 -5.29627316e-02 2.49963656e-01
2.21600875e-01 -6.63477659e-01 -2.12925255e-01 5.38554728e-01
6.02659285e-01 -1.18224287e+00 8.56304705e-01 -6.96591794e-01
-5.13695300e-01 -6.75797164e-01 -2.18734607e-01 -1.07136643e+00
-6.30459726e-01 -4.93376642e-01 -8.17242414e-02 8.19518864e-01
-1.12913162e-01 -6.06801569e-01 1.17673945e+00 6.80812180e-01
-3.97128195e-01 1.62267461e-02 -1.10242546e+00 -6.68704748e-01
-7.52765656e-01 -5.96293211e-01 2.12522328e-01 6.97803915e-01
-2.03877226e-01 2.21207425e-01 -1.95560470e-01 6.00209892e-01
5.55503845e-01 7.17018545e-02 1.28827405e+00 -7.25636005e-01
4.61474687e-01 -1.74141303e-01 -1.33221436e+00 -7.83046186e-01
2.30224267e-01 -2.51924068e-01 2.77874023e-01 -1.80966425e+00
2.17500255e-01 -1.88515887e-01 -8.15950409e-02 1.42830476e-01
1.51049122e-01 2.45664939e-01 -8.98620337e-02 -1.04926758e-01
-1.07796299e+00 2.33208910e-01 9.61109519e-01 3.26756477e-01
-2.22698987e-01 2.69860953e-01 1.20343743e-02 9.75624681e-01
9.42070007e-01 -4.72990960e-01 -3.99620652e-01 -8.82299840e-02
-2.36214399e-01 3.42277259e-01 2.64953285e-01 -1.25453782e+00
8.45039427e-01 -5.66680074e-01 1.69347629e-01 -1.23989940e+00
6.82345569e-01 -8.63080978e-01 7.00832188e-01 5.04845560e-01
4.81566221e-01 4.98848736e-01 2.37920329e-01 6.10514283e-01
-1.07854560e-01 -3.85196686e-01 5.05135536e-01 -4.02037710e-01
-1.74024653e+00 4.20870274e-01 -9.20636415e-01 -5.19514084e-01
1.71834409e+00 -7.80295193e-01 -1.61553964e-01 -8.70533526e-01
-7.80867636e-01 2.25861371e-01 8.26890945e-01 4.55620080e-01
4.75949734e-01 -1.17087233e+00 -5.24918020e-01 4.86288927e-02
-1.14155523e-02 -5.99298418e-01 4.44230169e-01 7.75374472e-01
-8.97200227e-01 1.05351341e+00 -4.24689382e-01 -6.89407766e-01
-1.65177608e+00 7.29579389e-01 2.28459224e-01 1.56415209e-01
-2.41544902e-01 5.20632029e-01 8.70213807e-02 5.83876111e-02
-3.47436130e-01 2.46630490e-01 -6.13183558e-01 -8.51696432e-02
3.73541921e-01 6.68735981e-01 -2.69824952e-01 -1.68753362e+00
-7.91959167e-01 8.36070836e-01 6.40814528e-02 -3.05844784e-01
9.62697446e-01 -6.42288983e-01 9.07168910e-02 5.08168995e-01
1.25426853e+00 2.78251976e-01 -1.13755774e+00 2.17634931e-01
-1.96053788e-01 -5.26537955e-01 3.94035250e-01 1.13634281e-01
-5.61145306e-01 6.25963271e-01 8.48644197e-01 6.61110207e-02
5.10517180e-01 9.19518471e-02 4.36692357e-01 4.60644990e-01
1.12757289e+00 -9.88058805e-01 -4.86786485e-01 8.43358099e-01
3.13545138e-01 -1.34331739e+00 2.51553029e-01 -2.46282622e-01
-6.77942455e-01 8.66869152e-01 7.03990400e-01 -1.13366239e-01
8.71020555e-01 3.07394415e-01 2.41493747e-01 -3.43486369e-01
-7.93697298e-01 -7.49301195e-01 4.60350402e-02 7.66333342e-01
1.03930384e-01 1.04937553e-01 1.42968476e-01 -5.07154882e-01
-2.35671282e-01 -4.13247906e-02 7.42345273e-01 1.14119875e+00
-8.06346178e-01 -6.76044881e-01 -3.34332585e-01 3.56893172e-03
-2.55164325e-01 3.27187330e-01 -1.96060747e-01 9.48784649e-01
1.80933252e-01 1.21702230e+00 3.19008619e-01 -3.83312345e-01
4.63339686e-01 -3.78364772e-01 3.07278633e-01 -5.88406324e-01
9.32443365e-02 -1.58872008e-01 3.51583362e-01 -8.42879713e-01
-4.54012960e-01 -1.18677580e+00 -1.32618821e+00 -6.92353308e-01
-3.43795210e-01 1.76405564e-01 1.33358073e+00 8.19368601e-01
3.32353622e-01 1.54002994e-01 7.53223777e-01 -1.23667192e+00
3.82904977e-01 -7.50664175e-01 -4.74024415e-01 1.03606157e-01
4.94778544e-01 -5.93569934e-01 -2.13737398e-01 -1.81639686e-01] | [6.030456066131592, 0.5716851949691772] |
e7c1ed7a-93ac-436d-b78f-cc5213fdfc5c | stock-trading-volume-prediction-with-dual | 2211.01762 | null | https://arxiv.org/abs/2211.01762v1 | https://arxiv.org/pdf/2211.01762v1.pdf | Stock Trading Volume Prediction with Dual-Process Meta-Learning | Volume prediction is one of the fundamental objectives in the Fintech area, which is helpful for many downstream tasks, e.g., algorithmic trading. Previous methods mostly learn a universal model for different stocks. However, this kind of practice omits the specific characteristics of individual stocks by applying the same set of parameters for different stocks. On the other hand, learning different models for each stock would face data sparsity or cold start problems for many stocks with small capitalization. To take advantage of the data scale and the various characteristics of individual stocks, we propose a dual-process meta-learning method that treats the prediction of each stock as one task under the meta-learning framework. Our method can model the common pattern behind different stocks with a meta-learner, while modeling the specific pattern for each stock across time spans with stock-dependent parameters. Furthermore, we propose to mine the pattern of each stock in the form of a latent variable which is then used for learning the parameters for the prediction module. This makes the prediction procedure aware of the data pattern. Extensive experiments on volume predictions show that our method can improve the performance of various baseline models. Further analyses testify the effectiveness of our proposed meta-learning framework. | ['Xu sun', 'Keiko Harimoto', 'Ruihan Bao', 'Zhiyuan Zhang', 'Wei Li', 'Ruibo Chen'] | 2022-10-11 | null | null | null | null | ['algorithmic-trading'] | ['time-series'] | [-5.32246530e-01 -3.11338902e-01 -7.47660875e-01 -2.15718940e-01
-3.85378599e-01 -4.89314795e-01 5.98269701e-01 4.70459387e-02
-2.40169182e-01 6.67389870e-01 2.37228706e-01 -1.36068135e-01
-3.80614661e-02 -1.12740517e+00 -7.69981444e-01 -6.93074703e-01
1.21644028e-01 6.26608193e-01 3.07068229e-01 1.53340315e-02
4.74601150e-01 1.77123263e-01 -1.32121396e+00 2.09015220e-01
8.14292431e-01 1.30307078e+00 4.30884629e-01 5.96229509e-02
-7.41836190e-01 9.80727315e-01 -4.03213948e-01 -4.75180209e-01
7.63622105e-01 -2.00288087e-01 -9.30519626e-02 3.10711563e-01
9.59120020e-02 -3.33496809e-01 -3.00013751e-01 1.06895852e+00
-1.15453109e-01 -8.63236785e-02 6.26137137e-01 -1.19537139e+00
-2.98681855e-01 1.07289112e+00 -9.82364357e-01 4.46194381e-01
-7.27497280e-01 2.75094926e-01 1.24712610e+00 -5.80034554e-01
1.29241720e-01 1.05331051e+00 4.51681525e-01 1.53243884e-01
-9.10911798e-01 -1.08158278e+00 7.64890075e-01 -2.09005103e-01
-9.19131041e-01 -7.11211935e-02 7.11870134e-01 -5.56191504e-01
2.63089687e-01 -1.97849661e-01 9.30742621e-01 6.83507681e-01
5.89286447e-01 1.05093336e+00 1.21048820e+00 2.55708575e-01
2.65457392e-01 3.76438826e-01 2.57812202e-01 3.77892405e-01
5.69057524e-01 -4.23362479e-02 -7.52858102e-01 -4.51373570e-02
9.31972742e-01 6.76481605e-01 1.75556913e-02 -3.68431717e-01
-1.13785529e+00 1.01707125e+00 3.72828878e-02 1.66776121e-01
-4.60189641e-01 2.49571845e-01 3.01299363e-01 4.90153015e-01
7.63683498e-01 2.58257300e-01 -9.68321562e-01 -8.59810859e-02
-1.09482491e+00 4.25311118e-01 9.79701042e-01 1.05284154e+00
1.04786181e+00 2.08771378e-01 -2.79519409e-01 5.49632549e-01
3.91088605e-01 3.28738302e-01 8.67299020e-01 -6.32309020e-01
8.78647923e-01 7.45433986e-01 3.09272379e-01 -7.85775900e-01
-3.11003119e-01 -9.60665107e-01 -7.92580843e-01 -2.78457645e-02
5.28648496e-01 -3.86478812e-01 -8.30959260e-01 1.59000134e+00
1.74385592e-01 6.06406808e-01 -5.96784763e-02 5.08098006e-01
2.29068384e-01 9.84235048e-01 3.24236415e-03 -5.05941510e-01
1.13498342e+00 -1.30632639e+00 -4.89275455e-01 -4.85196024e-01
5.56709230e-01 -5.62253177e-01 5.67508936e-01 4.96836394e-01
-9.60264087e-01 -5.44375122e-01 -8.24891925e-01 5.45909464e-01
-2.01562449e-01 1.00754850e-01 8.50664258e-01 3.62552762e-01
-4.44536567e-01 7.74925828e-01 -6.74354672e-01 3.73343110e-01
3.40798348e-01 2.97799468e-01 3.67186785e-01 3.96474600e-01
-9.60789800e-01 5.20328939e-01 6.31295979e-01 -2.46921517e-02
-6.97602510e-01 -1.27341354e+00 -3.46081406e-01 3.58373821e-01
6.44788146e-01 -4.78971273e-01 1.29807138e+00 -1.15949225e+00
-1.41621506e+00 3.28751147e-01 1.57565170e-03 -7.48791039e-01
7.69419372e-01 -3.46328855e-01 -3.31289679e-01 -4.92281526e-01
-4.65209782e-02 3.13195974e-01 1.13474166e+00 -1.10436344e+00
-1.28002977e+00 -3.06585848e-01 -1.16742633e-01 1.24881804e-01
-3.62527698e-01 -5.45512438e-01 -6.64513648e-01 -9.53931332e-01
1.00491583e-01 -8.88216496e-01 -2.27275193e-01 -5.46340525e-01
-2.08778709e-01 -3.08248606e-02 6.63966537e-01 -5.52764535e-01
1.41801167e+00 -1.91211843e+00 4.24172878e-02 3.45618606e-01
2.94131875e-01 -7.85245150e-02 7.30237886e-02 2.83404380e-01
9.07822773e-02 6.76063001e-02 3.71399783e-02 -3.39995265e-01
-1.12324759e-01 8.72936100e-02 -7.53323317e-01 1.35087922e-01
-1.12810545e-01 9.54201102e-01 -7.35887051e-01 -2.66765207e-01
-1.34122595e-02 -1.72433615e-01 -2.91712552e-01 3.51497442e-01
-5.45304060e-01 2.56603062e-01 -6.10913515e-01 4.38114464e-01
6.77087009e-01 -3.26732069e-01 2.16958001e-01 -6.74765036e-02
-3.59035224e-01 1.01870380e-01 -1.52394950e+00 1.25623131e+00
-7.59128392e-01 -8.98524001e-02 -1.89678416e-01 -7.44305372e-01
1.02686930e+00 -5.69083206e-02 5.49402475e-01 -5.49016237e-01
-4.38005105e-02 4.28742737e-01 1.62169442e-01 9.40263420e-02
4.13426369e-01 -4.74877119e-01 -4.07742634e-02 7.85772204e-01
1.91562925e-03 2.31532097e-01 1.52912542e-01 -5.28568551e-02
7.47712672e-01 -1.19957618e-01 2.15225071e-01 -2.21074909e-01
3.12908411e-01 -2.44473010e-01 1.08169603e+00 6.09714627e-01
-9.22191981e-03 9.65757072e-02 5.46507120e-01 -6.30344152e-01
-9.53799486e-01 -7.84582198e-01 2.23422319e-01 9.27901387e-01
1.86863065e-01 -3.35416198e-01 -1.79519907e-01 -6.78540468e-01
5.45533836e-01 5.42055964e-01 -6.62922144e-01 -4.87256348e-02
-3.32333684e-01 -1.17345786e+00 -1.88928977e-01 6.03213847e-01
6.05962992e-01 -1.00597703e+00 -3.05970430e-01 4.02423412e-01
3.80736381e-01 -7.57749021e-01 -5.90996385e-01 2.75018811e-01
-1.29474699e+00 -9.04204428e-01 -7.21605182e-01 -5.14476359e-01
1.48479268e-01 3.01165909e-01 1.04693484e+00 -6.60254359e-02
4.90570694e-01 8.04469213e-02 4.34290059e-02 -8.83390605e-01
-1.75562337e-01 4.81233001e-01 -3.72096337e-02 3.51342112e-01
2.74154156e-01 -6.08874202e-01 -5.79896510e-01 1.40778199e-01
-7.36152530e-01 5.50280251e-02 9.08021748e-01 7.48410523e-01
6.93456054e-01 5.19837141e-01 7.83501148e-01 -1.39799607e+00
6.13342404e-01 -8.20011973e-01 -1.05267406e+00 4.15256798e-01
-9.81486142e-01 3.73963416e-01 6.43737674e-01 -7.07899928e-01
-1.09876597e+00 -1.44712135e-01 5.40077329e-01 -7.24522233e-01
3.98505092e-01 6.40652716e-01 -2.80768186e-01 4.66199666e-01
-3.44924510e-01 5.02063453e-01 -6.45704120e-02 -5.55844069e-01
1.96829345e-03 2.58377105e-01 9.90004763e-02 -3.98446321e-01
1.22095895e+00 2.34550625e-01 -4.47216589e-04 -2.70297706e-01
-1.26534462e+00 -5.42579889e-01 -5.59902430e-01 3.31251160e-03
3.33143383e-01 -1.31179690e+00 -5.17942727e-01 8.33479941e-01
-5.99966109e-01 -4.46539193e-01 -3.48908216e-01 3.87179583e-01
-3.87113392e-01 -6.76986948e-02 -7.15740502e-01 -9.13632989e-01
-3.85308206e-01 -1.05449831e+00 8.68540764e-01 3.47774059e-01
1.36375502e-01 -1.43717051e+00 3.00077170e-01 1.83918148e-01
3.62412483e-01 -3.34746391e-01 9.11563337e-01 -9.58603740e-01
-1.05138683e+00 -5.91280274e-02 -5.07710278e-02 2.97516406e-01
3.70558262e-01 -2.42596656e-01 -7.88110316e-01 -1.34243935e-01
2.04660922e-01 -7.91246518e-02 1.19324458e+00 4.38017875e-01
1.31862807e+00 -4.17988122e-01 -1.55702561e-01 6.73735440e-01
1.55407560e+00 3.13817888e-01 3.19091499e-01 6.81263864e-01
7.37365127e-01 6.30544782e-01 7.00128078e-01 6.78036034e-01
3.30469042e-01 3.55067581e-01 1.78287655e-01 2.71516860e-01
5.49132705e-01 -6.13317251e-01 3.59341800e-01 9.22293186e-01
1.96252480e-01 1.82003733e-02 -6.73858404e-01 4.15047556e-01
-2.06696486e+00 -1.10017049e+00 2.60867357e-01 2.18878293e+00
7.97507346e-01 5.63834906e-01 4.05302078e-01 -1.24527425e-01
6.28937185e-01 4.84364271e-01 -9.24792051e-01 -6.03933893e-02
7.99786001e-02 -1.31876528e-01 9.89302158e-01 1.42322361e-01
-1.10351539e+00 8.79145324e-01 5.47252417e+00 9.52129781e-01
-1.15172935e+00 -1.83157906e-01 9.77323592e-01 -2.05121100e-01
-4.41191882e-01 8.20612684e-02 -1.41709280e+00 9.70453501e-01
9.61608648e-01 -5.92487693e-01 3.32200974e-01 9.22480047e-01
2.94041246e-01 1.88045744e-02 -9.33886588e-01 8.56357276e-01
-2.25682437e-01 -1.64028382e+00 3.23432177e-01 2.43648469e-01
9.06358361e-01 -1.06858388e-01 2.37724498e-01 6.46790087e-01
3.98770154e-01 -6.54508233e-01 8.59246552e-01 8.23108673e-01
1.63172781e-01 -1.00139272e+00 6.85073256e-01 7.25312114e-01
-1.54809141e+00 -4.65632021e-01 -5.37853360e-01 -4.38766144e-02
-1.14009671e-01 8.65019500e-01 -4.45683360e-01 4.94406641e-01
5.12499630e-01 1.02891314e+00 -4.83611435e-01 1.08576214e+00
1.14968628e-01 8.21345866e-01 -1.92301884e-01 1.99596137e-01
2.42244527e-01 -6.80244446e-01 9.33896676e-02 7.54558504e-01
5.12118876e-01 -2.24369705e-01 4.12204504e-01 7.50505626e-01
-4.51304257e-01 3.34440738e-01 -3.63500386e-01 -1.87687889e-01
5.31292021e-01 1.11005461e+00 -6.46967292e-01 -4.02036726e-01
-6.53364122e-01 4.56640601e-01 2.53499120e-01 1.60620794e-01
-6.35168910e-01 6.34218976e-02 4.45516676e-01 2.99867898e-01
6.22274578e-01 -6.26069009e-02 -6.33477390e-01 -1.32182074e+00
4.36989442e-02 -7.31239438e-01 4.54196125e-01 -1.02422513e-01
-1.64042342e+00 1.38399795e-01 -6.60364404e-02 -1.58152270e+00
-1.05964206e-01 -4.07357395e-01 -1.12772107e+00 9.34821188e-01
-1.75548184e+00 -9.52212453e-01 -3.09749022e-02 3.03149819e-01
7.96308398e-01 -6.46962047e-01 5.10827750e-02 -2.42168173e-01
-7.83622324e-01 2.61642843e-01 4.25149202e-01 1.07111707e-01
6.24059856e-01 -1.32306862e+00 3.33244056e-01 6.77473307e-01
1.14273459e-01 5.67687213e-01 4.14408028e-01 -9.82097447e-01
-1.30547333e+00 -1.54731333e+00 6.92615151e-01 -6.58445135e-02
1.23469043e+00 -9.05345678e-02 -1.18349195e+00 9.50212538e-01
1.05506241e-01 -3.12404007e-01 7.10143030e-01 3.14895988e-01
-2.81667054e-01 -6.06465816e-01 -7.02826321e-01 4.41863954e-01
5.57721138e-01 -5.95463440e-02 -5.64830482e-01 1.58544928e-01
8.05783272e-01 -2.35406637e-01 -8.95914674e-01 1.36479512e-01
3.65148753e-01 -9.22475338e-01 6.78922176e-01 -4.67782885e-01
4.80331153e-01 1.79433078e-02 1.53129280e-01 -1.48758876e+00
-4.11286652e-01 -4.30266768e-01 -6.35168433e-01 1.55913758e+00
5.09676516e-01 -6.99271798e-01 1.08274138e+00 6.52000248e-01
1.23125903e-01 -9.15921569e-01 -7.11130857e-01 -7.78170705e-01
3.45809579e-01 -2.13437200e-01 9.73057330e-01 7.55969107e-01
-3.55881631e-01 4.78408754e-01 -7.05464542e-01 1.15802184e-01
7.23840535e-01 8.76897335e-01 9.06058371e-01 -1.41553390e+00
-5.60440183e-01 -5.60984433e-01 2.83077043e-02 -1.11427128e+00
3.49172771e-01 -6.75765574e-01 -2.35979170e-01 -1.12794864e+00
3.97641718e-01 -7.68145800e-01 -5.68655908e-01 1.65166885e-01
-4.53480989e-01 -4.22781706e-01 4.75995570e-01 6.45066261e-01
-5.05450606e-01 6.93367600e-01 1.16631389e+00 -2.90490210e-01
-5.56117713e-01 6.17462933e-01 -1.03416491e+00 7.96495318e-01
1.09827781e+00 -3.99750024e-01 -6.40550733e-01 -2.08076447e-01
1.88700259e-01 2.55759507e-02 -6.98775575e-02 -7.96665490e-01
2.75606602e-01 -7.07660377e-01 5.22924900e-01 -8.62011552e-01
9.98625308e-02 -9.22533453e-01 2.30856270e-01 5.45075655e-01
-2.46893689e-01 3.07099521e-01 1.39103919e-01 9.75179017e-01
-3.14350337e-01 -2.36580983e-01 6.61052406e-01 -4.46165055e-01
-6.57806873e-01 8.20571899e-01 -2.12909207e-02 1.89407915e-01
1.06127536e+00 8.06666166e-02 -2.58040074e-02 -3.38972539e-01
-3.28839272e-01 7.30857849e-01 3.37703556e-01 3.85357946e-01
3.45117509e-01 -1.18247843e+00 -5.35241544e-01 1.00596972e-01
-6.32210746e-02 1.75731093e-01 2.73894340e-01 6.21286094e-01
-1.76895708e-02 4.02026713e-01 -1.34264350e-01 -3.58024091e-01
-6.15116894e-01 6.91771686e-01 4.55657661e-01 -9.84625995e-01
-4.97854799e-01 5.75104654e-01 6.18264318e-01 -3.70762870e-02
1.56884089e-01 -4.37828958e-01 -3.00613850e-01 5.46978295e-01
6.03978038e-01 2.04265565e-01 -7.44769424e-02 -1.17597260e-01
2.61350334e-01 4.79671180e-01 -4.27165866e-01 2.52189883e-03
1.75847101e+00 -1.82795182e-01 8.05208310e-02 1.18557811e+00
7.49496579e-01 1.09085612e-01 -1.60599983e+00 -5.89724422e-01
4.25785989e-01 -4.83565420e-01 3.83917056e-02 -4.58075464e-01
-1.71585906e+00 7.39085078e-01 1.45607248e-01 -1.55931935e-01
9.50934350e-01 -2.57592261e-01 9.41501915e-01 2.32127950e-01
4.38052386e-01 -1.13916183e+00 -6.75104675e-04 4.13932502e-01
4.51797336e-01 -1.21794951e+00 3.73193771e-02 -1.54942632e-01
-8.58264625e-01 1.02765501e+00 7.64650762e-01 -1.70951113e-01
1.13244462e+00 3.20813745e-01 3.90622988e-02 -2.76108477e-02
-1.25485778e+00 1.42189845e-01 2.08092064e-01 -6.16927445e-02
1.62756473e-01 9.25728530e-02 -1.04858868e-01 9.73945856e-01
-1.74832568e-01 1.20740384e-02 4.79618520e-01 8.83702517e-01
-6.23117268e-01 -1.26603353e+00 -1.69043362e-01 1.09383535e+00
-4.69107509e-01 -4.86183912e-02 -4.92923036e-02 8.14752281e-01
2.62114890e-02 3.66094947e-01 2.67851353e-01 -4.07467902e-01
4.39204574e-01 3.29234481e-01 5.09236343e-02 -8.94581854e-01
-8.78991663e-01 4.71290648e-01 -5.23653388e-01 -2.44071096e-01
-3.26478034e-01 -9.87282693e-01 -9.79765952e-01 -3.67719233e-01
-7.87202865e-02 2.35883340e-01 1.29838318e-01 9.94812489e-01
9.79513451e-02 4.35784042e-01 1.01041126e+00 -5.74168861e-01
-1.00118518e+00 -9.27653909e-01 -1.30954313e+00 2.95896024e-01
6.97398856e-02 -8.55735958e-01 -3.80813867e-01 -1.60440970e-02] | [4.4217753410339355, 4.237804889678955] |
2ea04f81-a234-4937-b867-82f68823c86f | bi-directional-object-context-prioritization | 2203.09416 | null | https://arxiv.org/abs/2203.09416v2 | https://arxiv.org/pdf/2203.09416v2.pdf | Bi-directional Object-context Prioritization Learning for Saliency Ranking | The saliency ranking task is recently proposed to study the visual behavior that humans would typically shift their attention over different objects of a scene based on their degrees of saliency. Existing approaches focus on learning either object-object or object-scene relations. Such a strategy follows the idea of object-based attention in Psychology, but it tends to favor those objects with strong semantics (e.g., humans), resulting in unrealistic saliency ranking. We observe that spatial attention works concurrently with object-based attention in the human visual recognition system. During the recognition process, the human spatial attention mechanism would move, engage, and disengage from region to region (i.e., context to context). This inspires us to model the region-level interactions, in addition to the object-level reasoning, for saliency ranking. To this end, we propose a novel bi-directional method to unify spatial attention and object-based attention for saliency ranking. Our model includes two novel modules: (1) a selective object saliency (SOS) module that models objectbased attention via inferring the semantic representation of the salient object, and (2) an object-context-object relation (OCOR) module that allocates saliency ranks to objects by jointly modeling the object-context and context-object interactions of the salient objects. Extensive experiments show that our approach outperforms existing state-of-theart methods. Our code and pretrained model are available at https://github.com/GrassBro/OCOR. | ['Rynson W. H. Lau', 'BaoCai Yin', 'Lin Du', 'Xin Yang', 'Ke Xu', 'Xin Tian'] | 2022-03-17 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Tian_Bi-Directional_Object-Context_Prioritization_Learning_for_Saliency_Ranking_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Tian_Bi-Directional_Object-Context_Prioritization_Learning_for_Saliency_Ranking_CVPR_2022_paper.pdf | cvpr-2022-1 | ['saliency-ranking'] | ['computer-vision'] | [ 2.48573348e-01 -9.15430859e-02 -2.82818019e-01 -4.22215134e-01
-3.06624472e-01 -1.23814300e-01 6.21093452e-01 3.76393676e-01
-3.06797296e-01 2.19295353e-01 4.29347515e-01 1.09565835e-02
-1.13471538e-01 -6.30978048e-01 -6.89500570e-01 -5.64418137e-01
1.68246806e-01 1.64690316e-01 8.03671420e-01 -2.27768272e-01
9.03159440e-01 3.63915294e-01 -1.88146317e+00 4.33882207e-01
9.54960346e-01 7.85296619e-01 8.04721534e-01 4.46930826e-01
-6.78882119e-04 1.06662059e+00 -3.16220641e-01 7.89008141e-02
-1.44854531e-01 -5.09874940e-01 -1.06091213e+00 8.37290362e-02
4.48550850e-01 -2.59852380e-01 -1.84990153e-01 1.27064240e+00
1.48219436e-01 3.72370064e-01 4.80346143e-01 -1.39109468e+00
-1.14839089e+00 5.10296106e-01 -6.70541644e-01 7.07010746e-01
2.58248210e-01 1.13744125e-01 1.16993487e+00 -1.30746424e+00
2.99779713e-01 1.42575037e+00 -1.02376141e-01 2.97534347e-01
-1.11696994e+00 -3.82208139e-01 6.95956290e-01 6.72729254e-01
-1.34665167e+00 -2.69265682e-01 1.17063260e+00 -4.46592331e-01
6.17635906e-01 4.99281347e-01 7.89759099e-01 8.23090613e-01
9.30964872e-02 1.23790348e+00 1.11616790e+00 -2.23335117e-01
2.60805279e-01 7.80871063e-02 4.64317560e-01 2.62030512e-01
3.02768081e-01 1.07615903e-01 -7.83542573e-01 5.25339469e-02
8.35676610e-01 3.63581151e-01 -1.85104325e-01 -6.21156454e-01
-1.43261564e+00 6.83283210e-01 1.15875483e+00 4.35096294e-01
-5.97452343e-01 2.91094452e-01 3.33971120e-02 -2.76934534e-01
2.41551474e-01 4.12360042e-01 -1.01448573e-01 4.75408465e-01
-8.21016490e-01 4.77198392e-01 7.41006956e-02 9.65820193e-01
8.71826589e-01 -1.10316209e-01 -6.78865492e-01 8.17482829e-01
6.08899474e-01 3.81600320e-01 5.17511189e-01 -6.00043237e-01
2.27127194e-01 7.69679546e-01 2.33048290e-01 -1.33469832e+00
-3.45202833e-01 -5.09780169e-01 -3.60081643e-01 1.05684370e-01
3.29630226e-02 5.93415678e-01 -9.78048086e-01 1.88659346e+00
4.59934711e-01 2.01284811e-01 -2.71777779e-01 1.53165376e+00
9.59234893e-01 3.61080438e-01 5.07671654e-01 1.10606402e-01
1.67218053e+00 -1.24643087e+00 -4.79925156e-01 -7.29323626e-01
1.28933862e-01 -5.48601210e-01 1.32465482e+00 -1.01955526e-01
-1.25182354e+00 -6.82219982e-01 -8.04466188e-01 -3.37659538e-01
-5.12817442e-01 -2.11346447e-02 6.23830616e-01 -1.59842953e-01
-1.08812594e+00 3.12247574e-01 -5.46810210e-01 -5.12940466e-01
7.67407477e-01 1.34155020e-01 1.48775727e-01 1.52853861e-01
-1.20009577e+00 1.08534157e+00 5.24769425e-01 6.06424995e-02
-1.23541903e+00 -4.83891308e-01 -8.98431897e-01 3.89281392e-01
5.43898463e-01 -5.63384950e-01 1.25742471e+00 -1.35982823e+00
-9.19932842e-01 1.08915663e+00 -5.45776546e-01 -9.76412073e-02
-1.69770941e-01 -4.46122766e-01 -1.82914615e-01 2.23193467e-01
4.47248578e-01 9.79806542e-01 9.61495578e-01 -1.59381318e+00
-8.10529411e-01 -4.26479578e-01 1.58717871e-01 4.45640653e-01
-5.61724640e-02 2.60748208e-01 -3.57901037e-01 -8.71759236e-01
4.92407709e-01 -7.10886955e-01 -1.11082762e-01 2.59345789e-02
-4.26207900e-01 -5.01782656e-01 6.60389721e-01 -4.63223815e-01
1.16392982e+00 -2.19117212e+00 3.39208901e-01 -1.37014538e-01
3.42823982e-01 4.15472724e-02 -1.78548157e-01 2.74080008e-01
-3.07874829e-01 -1.50166396e-02 -1.83531344e-01 -1.84553135e-02
-9.13260952e-02 -2.10347340e-01 -5.42678356e-01 3.35768580e-01
4.43203986e-01 1.20351815e+00 -1.50069118e+00 -6.13798797e-01
7.32283518e-02 2.57085204e-01 -5.46161532e-01 2.54771262e-01
-1.47316873e-01 3.33401680e-01 -5.71272433e-01 8.77631903e-01
6.05274379e-01 -5.34520149e-01 -1.97612159e-02 -1.51376650e-01
-1.55962750e-01 6.83435261e-01 -9.14748430e-01 1.56923628e+00
3.86681072e-02 4.59542781e-01 -1.79622889e-01 -8.06789637e-01
6.68724537e-01 -1.02996044e-02 1.77632160e-02 -8.63570690e-01
3.04734502e-02 1.83628738e-01 3.09356838e-01 -4.19763654e-01
6.61392808e-01 5.57558201e-02 6.46390952e-03 3.29212904e-01
-6.33407682e-02 8.26097578e-02 -5.21087982e-02 4.03320402e-01
6.33312285e-01 1.95335492e-01 4.69702125e-01 -7.12936521e-01
3.95747364e-01 -4.49367240e-02 5.15164614e-01 6.88632667e-01
-6.14998043e-01 5.75410008e-01 3.50443125e-01 -3.16245139e-01
-6.24448717e-01 -1.18624222e+00 5.31965904e-02 1.74052858e+00
1.04472780e+00 5.08646853e-02 -6.08020544e-01 -5.85429251e-01
-8.58556759e-03 9.62565482e-01 -9.39380705e-01 -4.53052282e-01
-5.25642276e-01 -2.94409901e-01 -2.87141144e-01 7.30133832e-01
3.65680277e-01 -1.81203210e+00 -1.10146534e+00 -1.48360729e-01
-1.84345946e-01 -5.49910784e-01 -8.11579585e-01 -1.61408503e-02
-7.95864582e-01 -1.04474175e+00 -5.95422566e-01 -1.00489593e+00
9.11765516e-01 8.62036884e-01 1.20402920e+00 4.40192133e-01
-5.61497509e-02 3.82032365e-01 -3.54773611e-01 -6.41530156e-01
2.79253602e-01 -2.10934635e-02 -2.19476949e-02 1.95484057e-01
5.18867910e-01 -3.68361473e-01 -9.88351464e-01 5.03974199e-01
-1.04571927e+00 4.79392856e-01 7.29380369e-01 6.12267196e-01
5.38403869e-01 -3.85246813e-01 7.03638673e-01 -5.02126276e-01
3.54538321e-01 -7.18749106e-01 -4.96862113e-01 3.23546588e-01
-2.57426083e-01 -1.37200743e-01 -1.47758587e-03 -5.04558861e-01
-9.54526901e-01 -1.18602015e-01 4.94061291e-01 -4.61694747e-01
-1.81248635e-01 4.74786490e-01 -2.44500250e-01 1.69918552e-01
4.01856422e-01 4.62033153e-01 -4.54043031e-01 -1.73240334e-01
2.84383655e-01 3.01839650e-01 4.27446991e-01 -4.38779056e-01
6.99129462e-01 5.68751991e-01 -2.53805876e-01 -4.71064150e-01
-1.28825569e+00 -6.02855623e-01 -6.46473706e-01 -3.28964978e-01
9.63546455e-01 -9.13659215e-01 -5.87627769e-01 4.17250961e-01
-1.23581481e+00 -2.95352906e-01 -2.56123066e-01 3.85712981e-01
-5.14915228e-01 -4.69940417e-02 -2.88456291e-01 -7.56260514e-01
8.24556628e-04 -1.07257354e+00 1.33604646e+00 5.71110129e-01
-2.97320873e-01 -6.78131461e-01 -3.02202314e-01 1.80235341e-01
4.02647883e-01 -1.42340913e-01 6.66085720e-01 -6.02998137e-01
-1.08265841e+00 3.07345212e-01 -6.51374280e-01 -3.37165684e-01
1.96771458e-01 -3.62447292e-01 -9.94857788e-01 -1.37261167e-01
1.68474406e-01 -2.95701444e-01 1.13737679e+00 4.13696498e-01
1.30856657e+00 -3.29510719e-01 -5.46439469e-01 2.82129735e-01
1.19400525e+00 1.78526342e-01 7.43552089e-01 2.90537447e-01
8.58112097e-01 8.43461752e-01 1.02731478e+00 1.48837268e-01
7.05594301e-01 7.37790167e-01 7.04035163e-01 -1.26668781e-01
-1.03679717e-01 -2.89465010e-01 1.37509838e-01 3.51401031e-01
1.73464641e-01 3.59491035e-02 -9.11152780e-01 1.01167452e+00
-2.10113192e+00 -1.07359254e+00 -1.45035192e-01 2.18399191e+00
8.72679889e-01 7.52439201e-02 1.79120183e-01 -1.22368187e-01
9.24121916e-01 2.83109039e-01 -7.12430894e-01 -1.82200745e-01
9.64514762e-02 -3.61704469e-01 -4.81580831e-02 5.08962512e-01
-1.11106658e+00 1.33640766e+00 5.51678038e+00 5.67162693e-01
-1.18384862e+00 1.65497288e-01 5.26935637e-01 -1.74955636e-01
-4.29239720e-01 1.56966910e-01 -7.24147439e-01 5.13661087e-01
1.95807248e-01 -2.64883161e-01 3.65451574e-01 9.66979980e-01
-2.57557817e-03 -4.38761145e-01 -1.21707952e+00 7.34528184e-01
3.38563621e-01 -9.71777439e-01 2.78484434e-01 -1.54527381e-01
4.45541680e-01 -1.60494894e-01 2.70450860e-01 2.09618017e-01
1.17956968e-02 -9.41987097e-01 1.35432220e+00 7.31905878e-01
2.69206673e-01 -3.00970584e-01 3.21982354e-01 2.89366841e-01
-1.23652685e+00 -1.30227476e-01 -3.43474686e-01 -1.42994106e-01
-3.97493280e-02 3.94790798e-01 -3.86248589e-01 6.49611428e-02
9.03380334e-01 7.75039792e-01 -7.97972858e-01 1.18531287e+00
-5.92129469e-01 2.56221861e-01 2.93395817e-01 -2.23371670e-01
1.95699349e-01 1.39735162e-01 8.28188956e-01 8.85616422e-01
-1.88121144e-02 3.61032963e-01 2.02952683e-01 1.33612895e+00
2.49140278e-01 -4.63469587e-02 -2.61561334e-01 2.61541337e-01
5.15915215e-01 1.22543192e+00 -9.42349315e-01 -6.00380003e-01
-3.15612257e-01 9.33297873e-01 6.12199545e-01 5.54890931e-01
-9.82283056e-01 -2.51260668e-01 4.82048571e-01 2.40316197e-01
5.63689768e-01 -3.75760831e-02 -4.72998470e-01 -1.04189861e+00
4.75600921e-02 -5.66649497e-01 3.51423651e-01 -1.23177481e+00
-1.04780328e+00 3.98937106e-01 1.21415459e-01 -1.13792384e+00
4.36484754e-01 -3.78020048e-01 -8.13336313e-01 1.08602750e+00
-1.79228425e+00 -1.23054266e+00 -5.14949620e-01 3.73301327e-01
8.14671159e-01 2.65517682e-01 2.13029951e-01 -2.08814889e-01
-3.73252869e-01 2.08872721e-01 -7.43699193e-01 -5.13373315e-02
4.41290259e-01 -1.13371587e+00 1.59498438e-01 8.72678578e-01
1.16254792e-01 9.28737044e-01 5.62824547e-01 -8.66188586e-01
-1.09823263e+00 -9.92035091e-01 1.20340383e+00 -7.46169567e-01
5.20264447e-01 -4.00881588e-01 -1.27737784e+00 6.45728052e-01
1.66130558e-01 1.56111065e-02 3.89783412e-01 2.31554419e-01
-3.17282587e-01 4.83013922e-03 -7.89462924e-01 9.80969489e-01
1.24804688e+00 -5.73631465e-01 -1.13539112e+00 2.04375967e-01
8.82272243e-01 -2.56844163e-01 -1.14439771e-01 4.26289290e-01
3.64241004e-01 -1.09167755e+00 1.16231167e+00 -6.26380324e-01
6.37809277e-01 -7.67780840e-01 -1.61082953e-01 -1.11190295e+00
-1.06474817e+00 2.38864589e-02 -3.33636969e-01 1.02760351e+00
2.08857924e-01 -4.41356093e-01 3.31097096e-01 4.13084865e-01
-2.75032133e-01 -7.50618279e-01 -6.52028322e-01 -5.67187428e-01
-3.12032342e-01 -6.76476955e-02 6.53001666e-01 9.52316761e-01
5.05867824e-02 4.80477989e-01 -6.60382658e-02 5.30762196e-01
4.85226959e-01 5.80816507e-01 2.48218164e-01 -1.21454215e+00
-2.36831931e-03 -8.41531813e-01 -2.71517932e-01 -1.01208544e+00
7.28674904e-02 -9.41791415e-01 3.94119710e-01 -1.57897961e+00
7.11979449e-01 -3.94614935e-01 -8.51323545e-01 6.80573583e-01
-9.07127023e-01 1.34958131e-02 3.73646200e-01 4.93447393e-01
-1.14197373e+00 6.90094054e-01 1.43197179e+00 -1.24468535e-01
-2.78742909e-01 -3.35213691e-01 -1.20002079e+00 8.43934596e-01
7.93718338e-01 -3.90934169e-01 -4.26508337e-01 -5.68535328e-01
3.30205150e-02 -2.71331400e-01 9.81312096e-01 -9.02213395e-01
3.07654947e-01 -7.04195857e-01 5.43364882e-01 -7.05634892e-01
1.86507925e-01 -6.80183232e-01 -5.02447963e-01 4.87025499e-01
-5.92330456e-01 -9.25123245e-02 1.24403201e-01 4.05827820e-01
-2.09316477e-01 -1.05632186e-01 8.52615476e-01 -2.20387876e-01
-1.05218577e+00 2.27579907e-01 -1.40120074e-01 3.84557210e-02
1.06717813e+00 -1.44045785e-01 -5.17538905e-01 -1.64147049e-01
-6.04642868e-01 4.10168469e-01 4.77995336e-01 8.08240891e-01
9.50847924e-01 -1.62241733e+00 -6.13957524e-01 9.02805626e-02
4.33971941e-01 -2.63912380e-02 2.86061764e-01 9.74539578e-01
1.06259529e-02 4.72497493e-01 -1.36771172e-01 -7.59134591e-01
-1.03561342e+00 1.09227514e+00 2.33144119e-01 2.40666032e-01
-1.91482082e-01 1.07917142e+00 1.06043911e+00 -4.63276356e-02
7.10822344e-02 -4.51185435e-01 -5.20704865e-01 -9.08667818e-02
7.43793070e-01 2.10055187e-01 -4.04782265e-01 -8.67591500e-01
-7.49708712e-01 5.56406498e-01 -1.30683839e-01 5.18447571e-02
9.59090471e-01 -1.31177023e-01 -4.83909756e-01 7.36570776e-01
6.72700047e-01 -1.59480944e-01 -1.24690652e+00 -3.68496925e-01
2.06236839e-01 -8.14972043e-01 -2.32224520e-02 -8.08610678e-01
-7.86450446e-01 8.65376770e-01 4.00337756e-01 2.49229729e-01
1.26705194e+00 6.19725347e-01 3.91810536e-01 -1.08463457e-02
4.01927024e-01 -9.70072746e-01 5.52496552e-01 5.24764717e-01
1.41984713e+00 -1.43349719e+00 8.81291479e-02 -5.76637447e-01
-8.60059738e-01 4.12685633e-01 1.18053317e+00 -3.17541629e-01
6.18242145e-01 -4.66819853e-01 -1.22872606e-01 -4.89743114e-01
-8.00784528e-01 -5.77821672e-01 7.64416575e-01 3.99491668e-01
4.46321487e-01 1.70682907e-01 -1.55123860e-01 7.14078248e-01
1.73699170e-01 -1.51145190e-01 1.32083938e-01 1.06668544e+00
-8.77151608e-01 -5.16349554e-01 -5.26305377e-01 3.11765075e-01
2.70090345e-02 -4.46572840e-01 -4.89653349e-01 3.74566436e-01
1.45153299e-01 8.37949455e-01 3.36221039e-01 -4.05215025e-02
2.23756477e-01 -1.40890926e-01 2.13970363e-01 -9.20182765e-01
-4.54167694e-01 3.82891484e-02 -5.41367769e-01 -7.18032718e-01
-5.69844782e-01 -7.96180487e-01 -1.42641127e+00 2.48452082e-01
-2.74222285e-01 -5.83687760e-02 2.52352297e-01 8.48725140e-01
5.01652062e-01 7.03662038e-01 5.81935346e-01 -1.07318449e+00
-1.08958129e-02 -8.21702123e-01 -5.12845218e-01 6.70785785e-01
4.79970187e-01 -1.01409554e+00 -2.19309509e-01 -4.55825515e-02] | [9.941141128540039, 0.08417186141014099] |
cdd229fc-756c-4446-9a28-66426e27f5f7 | point-based-value-iteration-for-neuro | 2306.17639 | null | https://arxiv.org/abs/2306.17639v1 | https://arxiv.org/pdf/2306.17639v1.pdf | Point-based Value Iteration for Neuro-Symbolic POMDPs | Neuro-symbolic artificial intelligence is an emerging area that combines traditional symbolic techniques with neural networks. In this paper, we consider its application to sequential decision making under uncertainty. We introduce neuro-symbolic partially observable Markov decision processes (NS-POMDPs), which model an agent that perceives a continuous-state environment using a neural network and makes decisions symbolically, and study the problem of optimising discounted cumulative rewards. This requires functions over continuous-state beliefs, for which we propose a novel piecewise linear and convex representation (P-PWLC) in terms of polyhedra covering the continuous-state space and value vectors, and extend Bellman backups to this representation. We prove the convexity and continuity of value functions and present two value iteration algorithms that ensure finite representability by exploiting the underlying structure of the continuous-state model and the neural perception mechanism. The first is a classical (exact) value iteration algorithm extending $\alpha$-functions of Porta et al (2006) to the P-PWLC representation for continuous-state spaces. The second is a point-based (approximate) method called NS-HSVI, which uses the P-PWLC representation and belief-value induced functions to approximate value functions from below and above for two types of beliefs, particle-based and region-based. Using a prototype implementation, we show the practical applicability of our approach on two case studies that employ (trained) ReLU neural networks as perception functions, dynamic car parking and an aircraft collision avoidance system, by synthesising (approximately) optimal strategies. An experimental comparison with the finite-state POMDP solver SARSOP demonstrates that NS-HSVI is more robust to particle disturbances. | ['Marta Kwiatkowska', 'David Parker', 'Gethin Norman', 'Gabriel Santos', 'Rui Yan'] | 2023-06-30 | null | null | null | null | ['decision-making-under-uncertainty', 'decision-making-under-uncertainty', 'decision-making'] | ['medical', 'reasoning', 'reasoning'] | [ 2.20030650e-01 7.06577659e-01 -2.17740938e-01 -1.85692832e-01
-2.89531559e-01 -5.51135600e-01 6.94815218e-01 2.33543113e-01
-6.79471672e-01 1.26191747e+00 -1.45555034e-01 -4.66794908e-01
-7.15559423e-01 -9.43931103e-01 -8.31710637e-01 -7.88479924e-01
-4.62640822e-01 1.09530938e+00 2.96951026e-01 -5.75637281e-01
8.54065120e-02 6.55211270e-01 -1.64279437e+00 1.24917530e-01
4.75465953e-01 1.24802601e+00 1.24856159e-01 5.98192215e-01
1.86943207e-02 9.86791849e-01 -4.83227432e-01 1.09294057e-01
4.07178015e-01 -2.09006235e-01 -8.51886928e-01 -2.00648934e-01
-6.97358251e-01 -2.76818514e-01 1.38883684e-02 1.19942474e+00
3.15235518e-02 7.52722561e-01 6.08139634e-01 -1.67055583e+00
-4.06748533e-01 7.46615767e-01 2.16004193e-01 -1.53168157e-01
3.65601391e-01 2.94396102e-01 3.66458952e-01 -1.47828013e-01
6.02455914e-01 1.84309363e+00 5.23962438e-01 6.04141057e-01
-1.29290521e+00 1.99693535e-02 3.17016810e-01 3.66054833e-01
-1.25808525e+00 3.69312204e-02 3.24006855e-01 -1.05616473e-01
1.52633369e+00 3.89338344e-01 1.04892790e+00 9.21003401e-01
7.27990389e-01 7.87149906e-01 1.59931934e+00 -3.34868580e-01
1.18091691e+00 -2.08908752e-01 -8.33868757e-02 4.30040240e-01
1.07075743e-01 8.52848887e-01 3.16759124e-02 -4.45424050e-01
8.66044402e-01 9.10457000e-02 4.53692190e-02 -5.84224999e-01
-1.17257071e+00 8.24310124e-01 2.41739511e-01 5.83890267e-02
-8.64408791e-01 4.27847981e-01 2.82291293e-01 3.39377880e-01
-4.27833647e-02 4.24651444e-01 -3.02434653e-01 -2.73983777e-01
-5.81372619e-01 8.43549848e-01 1.34657836e+00 8.05998266e-01
4.56836820e-01 3.60251248e-01 -3.52139294e-01 5.44153899e-03
5.94799399e-01 8.12486410e-01 3.67606848e-01 -1.63084614e+00
2.95313839e-02 2.72754699e-01 8.24549079e-01 -5.47649920e-01
-5.19827187e-01 6.27001747e-02 -4.18231189e-01 9.08857226e-01
2.41210833e-01 -3.52768481e-01 -1.06436551e+00 1.68827593e+00
3.36779743e-01 2.19952628e-01 7.19312251e-01 8.86846006e-01
-1.46076143e-01 9.51332211e-01 -1.21277019e-01 -5.32321393e-01
8.42029393e-01 -6.62132680e-01 -7.98458993e-01 8.09322968e-02
5.51190302e-02 2.87370324e-01 4.08092648e-01 7.56607413e-01
-1.46196759e+00 -2.78121024e-01 -1.25092304e+00 5.51827967e-01
-3.89898747e-01 -6.58283353e-01 5.06549954e-01 4.78452533e-01
-1.39535010e+00 1.17973280e+00 -1.29903889e+00 -1.12353466e-01
-8.69515091e-02 7.52536893e-01 1.92785293e-01 2.16457427e-01
-1.45271778e+00 1.35211337e+00 1.06242228e+00 1.98409945e-01
-1.47099924e+00 8.55614692e-02 -8.44572008e-01 2.46427029e-01
8.52052212e-01 -4.43584681e-01 1.66829360e+00 -7.62804508e-01
-2.18409657e+00 1.72518101e-02 1.74443677e-01 -1.09728968e+00
4.10643280e-01 2.62556881e-01 -1.77798644e-01 1.69346720e-01
-2.85179138e-01 6.76706493e-01 7.23738790e-01 -1.50293326e+00
-7.32670665e-01 -2.06327781e-01 4.21586961e-01 4.27602887e-01
5.83453774e-01 -2.65607804e-01 3.09932798e-01 4.99558412e-02
2.46343449e-01 -1.14675713e+00 -8.56855989e-01 -3.70546669e-01
-1.17724068e-01 -1.96220934e-01 3.35823238e-01 -2.14384913e-01
6.87353134e-01 -1.57201529e+00 5.17448366e-01 6.44571841e-01
-2.18692973e-01 1.04612848e-02 8.25783983e-03 5.54696143e-01
1.20282203e-01 -1.81251988e-01 -6.43271625e-01 -2.45873615e-01
5.26999414e-01 1.11849940e+00 -5.57870567e-01 3.60074967e-01
1.07641332e-01 8.28299165e-01 -1.14069140e+00 -2.71969259e-01
2.59909928e-01 2.03785405e-01 -1.80978030e-01 -2.24635094e-01
-7.80931890e-01 1.51015833e-01 -5.20961344e-01 5.92731118e-01
3.86175275e-01 5.78048885e-01 3.86470467e-01 6.69961095e-01
-5.19344091e-01 -1.49015352e-01 -1.71227813e+00 1.32707274e+00
-1.14860021e-01 4.68085147e-02 4.75571126e-01 -9.60481882e-01
8.29071999e-01 4.27844077e-01 3.41466099e-01 -4.29930806e-01
5.87522805e-01 1.38539359e-01 -9.65478048e-02 -2.15176959e-02
6.53988838e-01 -5.85674822e-01 -3.94422978e-01 2.64005154e-01
-8.37764144e-03 -5.34659147e-01 1.09220952e-01 -3.14171493e-01
1.11136293e+00 5.56309521e-01 4.87836212e-01 -4.60739255e-01
4.99895245e-01 3.89505297e-01 7.73662031e-01 1.02385175e+00
-5.99911332e-01 1.65352240e-01 8.54666352e-01 -4.24235344e-01
-7.17233062e-01 -1.22815931e+00 5.22698537e-02 8.92981827e-01
3.68382961e-01 3.47620726e-01 -9.04684603e-01 -1.23705208e-01
1.90702200e-01 1.48403013e+00 -7.69292295e-01 -1.37991294e-01
-3.86703670e-01 -4.06227767e-01 5.10627866e-01 5.54902971e-01
2.84730971e-01 -1.45265317e+00 -1.24200773e+00 6.04008317e-01
2.17766538e-01 -5.42066038e-01 3.81110072e-01 6.93488479e-01
-8.18789661e-01 -6.16022229e-01 -3.75603437e-01 -3.12857330e-01
3.71521741e-01 -4.66383636e-01 6.70500040e-01 -5.29998541e-01
5.23813009e-01 7.78190553e-01 -7.51160160e-02 -8.97698104e-01
-6.84911549e-01 -6.85657442e-01 5.60743272e-01 -1.16813652e-01
-3.05048171e-02 -4.21098679e-01 6.93151429e-02 1.89485878e-01
-9.63855565e-01 -2.83428654e-03 3.61544371e-01 7.71524370e-01
9.78992999e-01 2.86985844e-01 8.27210620e-02 -1.71682239e-01
1.09532928e+00 -4.38240707e-01 -1.07874739e+00 1.42346293e-01
-5.48069656e-01 5.05974054e-01 4.45859492e-01 -3.83126825e-01
-1.24481857e+00 1.51866555e-01 8.94821435e-02 -7.26628244e-01
-2.86166131e-01 6.68043017e-01 -4.91873473e-02 -5.15686814e-04
5.02817690e-01 3.28552425e-01 3.47324491e-01 1.32367209e-01
2.99704343e-01 2.74712235e-01 7.07601547e-01 -9.84253526e-01
3.04941207e-01 7.16559529e-01 5.31951666e-01 -3.66922110e-01
-2.42555723e-01 9.94547382e-02 -4.04334933e-01 -3.86995018e-01
8.35179567e-01 -3.21239769e-01 -1.48824692e+00 1.71416193e-01
-1.20749569e+00 -5.77342808e-01 -1.10599184e+00 4.94910181e-01
-1.48350537e+00 3.42895314e-02 -5.35505891e-01 -1.58610094e+00
1.07341766e-01 -1.35665023e+00 6.48292780e-01 9.45459306e-02
4.07517292e-02 -7.94709504e-01 4.18399602e-01 -3.38711113e-01
2.82624483e-01 7.57823765e-01 6.19282424e-01 -7.77311563e-01
-3.99077922e-01 3.59491333e-02 3.17640454e-01 2.51638085e-01
-7.57397294e-01 -4.35994059e-01 -7.20392227e-01 -3.01436812e-01
5.28492987e-01 -1.81205556e-01 4.20632571e-01 5.64433396e-01
6.00970864e-01 -5.46967506e-01 -3.72498453e-01 7.42579475e-02
1.45541787e+00 1.01116502e+00 5.04101932e-01 6.63843453e-01
2.33603884e-02 6.24773443e-01 9.53181326e-01 6.17733002e-01
4.58716899e-01 4.50510561e-01 1.11117005e+00 6.42284274e-01
7.86907017e-01 -2.20755115e-02 7.57803857e-01 2.23717228e-01
-6.74800396e-01 -6.84934705e-02 -1.06926370e+00 4.27645504e-01
-2.47414327e+00 -1.28686261e+00 2.18664601e-01 2.23429656e+00
6.12265885e-01 4.42786425e-01 8.50107148e-02 1.02119379e-01
8.23447406e-01 -3.10353994e-01 -8.45383823e-01 -1.29850388e+00
1.80454254e-01 1.26185820e-01 7.36181617e-01 8.00297797e-01
-7.93774545e-01 7.42005885e-01 6.18085241e+00 5.94601452e-01
-7.62649298e-01 2.42822260e-01 2.40757585e-01 -2.90142119e-01
-9.05363038e-02 3.70519701e-03 -7.41197050e-01 2.62423277e-01
1.66975844e+00 -2.66292423e-01 1.08317828e+00 1.11682856e+00
2.67498404e-01 -3.86485904e-01 -1.14292669e+00 5.92200160e-01
-3.46188426e-01 -1.27013922e+00 -3.51642847e-01 1.15486555e-01
6.71252966e-01 1.18843466e-02 -4.85912748e-02 7.85181761e-01
1.01091731e+00 -1.07834446e+00 1.54126441e+00 7.70342827e-01
3.97104949e-01 -1.26492715e+00 9.33816731e-01 8.17498803e-01
-9.64586020e-01 -6.65810585e-01 -4.36809927e-01 -6.10843837e-01
5.70697427e-01 -1.75307050e-01 -7.72629321e-01 6.23022020e-01
4.48255122e-01 1.12802252e-01 3.86165231e-01 6.21553898e-01
-1.52101368e-01 1.70402244e-01 -7.26468384e-01 -5.01233935e-01
9.05512214e-01 -3.27247560e-01 9.61665273e-01 8.18018436e-01
3.11038673e-01 3.25084627e-01 4.17227536e-01 1.23728812e+00
8.83489788e-01 -6.92877710e-01 -5.39525867e-01 2.85104308e-02
3.22610080e-01 6.49549544e-01 -9.94522989e-01 -3.43300939e-01
2.15876013e-01 6.14959478e-01 -9.30393953e-03 3.28860611e-01
-9.00004685e-01 -1.09617591e-01 7.50898480e-01 -3.40806693e-01
3.07033509e-01 -4.37262952e-01 -9.47238654e-02 -6.09155476e-01
-5.22792935e-01 -6.14203036e-01 1.13439530e-01 -8.71960104e-01
-7.47093201e-01 6.04525387e-01 9.04058516e-01 -1.06204844e+00
-9.45827365e-01 -6.93455994e-01 -3.09421360e-01 8.30325842e-01
-1.16134632e+00 -8.52805316e-01 3.93406928e-01 6.54217064e-01
1.86284333e-01 2.59236363e-03 9.30609286e-01 -7.38829613e-01
-2.38040984e-01 -3.60818863e-01 4.62488264e-01 -6.08521461e-01
-4.22812849e-01 -1.44452727e+00 3.03203076e-01 5.26994824e-01
-4.67229754e-01 3.60947043e-01 1.07948470e+00 -7.93968856e-01
-1.53974462e+00 -8.50371957e-01 2.76170194e-01 -4.67636317e-01
6.56124532e-01 9.01120976e-02 -6.64392292e-01 8.83055866e-01
1.26854077e-01 -2.54311085e-01 -2.97090523e-02 -4.65071917e-01
5.38075268e-01 3.53682727e-01 -1.77335417e+00 7.28164256e-01
9.10530150e-01 -2.81091174e-03 -1.13132012e+00 -3.57809626e-02
8.26932311e-01 -7.95632720e-01 -7.68198669e-01 3.38866591e-01
3.65608901e-01 -9.64311182e-01 9.46399331e-01 -6.36163890e-01
-8.06818977e-02 -4.85779971e-01 -3.77263397e-01 -1.41375923e+00
-3.91866148e-01 -6.83452606e-01 -4.35160130e-01 5.19374609e-01
7.17527494e-02 -1.01668549e+00 5.24766624e-01 8.29790592e-01
-4.76350158e-01 -1.08822489e+00 -1.64774108e+00 -1.09923518e+00
1.00787356e-02 -8.51987720e-01 8.48589778e-01 4.67303842e-01
3.11610430e-01 -4.01830405e-01 1.77746356e-01 2.39913180e-01
7.61978090e-01 -1.08001955e-01 -5.00838459e-02 -1.07435596e+00
-5.76043010e-01 -2.45815262e-01 -3.76654834e-01 -4.84363407e-01
2.89372087e-01 -6.19853616e-01 6.22396290e-01 -1.56545234e+00
-4.29561764e-01 -3.20557564e-01 -6.26445174e-01 7.35390842e-01
7.70369172e-01 -5.79244673e-01 6.27118230e-01 1.09377049e-01
-6.18821621e-01 6.60666108e-01 1.23280418e+00 1.07708713e-02
-5.38588405e-01 8.32172036e-02 -1.98240474e-01 1.08273506e+00
7.57570922e-01 -4.37851340e-01 -5.78060508e-01 9.22961459e-02
3.91488671e-01 7.47456610e-01 4.65108961e-01 -1.39051938e+00
3.01420748e-01 -8.53255630e-01 5.72856776e-02 -6.51422083e-01
8.85403097e-01 -9.66391563e-01 4.65692163e-01 9.50769067e-01
-3.18590701e-01 3.05465162e-01 4.79065657e-01 8.65778804e-01
3.17116491e-02 -6.44867122e-01 5.64459741e-01 -4.47702914e-01
-9.61730063e-01 -2.05320910e-01 -9.77151155e-01 -4.54321384e-01
1.36891580e+00 -3.40415388e-01 -7.04995170e-02 -3.69639039e-01
-1.30498445e+00 3.22855681e-01 3.87826562e-01 -1.15413152e-01
7.74820685e-01 -1.07692051e+00 -1.83493912e-01 1.55610871e-02
-5.70492148e-01 2.05136061e-01 1.96898676e-04 6.92408264e-01
-5.40221155e-01 6.20392144e-01 -6.56613529e-01 -4.74765807e-01
-6.24326766e-01 8.01834643e-01 5.29848874e-01 -3.84613574e-01
-2.73311734e-01 4.87614870e-01 -4.69864190e-01 -5.71002781e-01
1.85426384e-01 -9.66322541e-01 -3.28256041e-01 -4.93072048e-02
4.07220930e-01 7.88533866e-01 -3.67965311e-01 -5.11275530e-01
-1.98952794e-01 -1.71438470e-01 5.79196155e-01 -9.85403955e-01
1.29904127e+00 -2.98638642e-02 -1.34165123e-01 7.07158506e-01
2.52798796e-01 -9.07700539e-01 -1.70975637e+00 1.25692531e-01
-4.04018983e-02 1.57596484e-01 1.07574672e-01 -8.70049536e-01
-3.17624331e-01 4.69486743e-01 6.77901804e-01 5.44911444e-01
7.62892783e-01 -4.06707764e-01 3.68647516e-01 1.01527071e+00
1.08340633e+00 -1.32440507e+00 -6.17311478e-01 1.09165180e+00
1.01843476e+00 -5.38577497e-01 -2.57557929e-01 9.63398218e-02
-9.77645814e-01 1.03460371e+00 2.56319731e-01 -3.20911556e-01
3.32096785e-01 5.11126399e-01 -3.12910497e-01 8.19464996e-02
-9.30502355e-01 -4.59275573e-01 -2.73727030e-01 8.95841897e-01
-7.78019965e-01 5.05344927e-01 -6.25295490e-02 8.04318249e-01
-1.17212176e-01 3.25468510e-01 7.34751046e-01 1.56366289e+00
-7.52188981e-01 -7.29791999e-01 -1.09183097e+00 1.04271650e-01
-1.55414362e-02 4.03165460e-01 -2.44156402e-02 8.13240707e-01
2.70824879e-01 9.56559360e-01 2.16856480e-01 -2.48617515e-01
4.43968475e-01 1.83825806e-01 6.27545893e-01 -5.67883790e-01
-6.14291906e-01 -2.91451961e-01 1.47972167e-01 -8.46714556e-01
-3.13194513e-01 -9.97121453e-01 -1.94549263e+00 -3.54626447e-01
1.31215349e-01 1.04402728e-01 9.71190095e-01 1.21587837e+00
5.57081550e-02 5.81435263e-01 6.69560805e-02 -1.56958687e+00
-1.33067393e+00 -5.22791207e-01 -7.61720359e-01 -1.84559315e-01
1.82071969e-01 -1.00634265e+00 -2.09865212e-01 -3.20576906e-01] | [4.36722469329834, 2.1570072174072266] |
27602f8b-4925-4e67-b15d-e1b471cc70db | lightea-a-scalable-robust-and-interpretable | 2210.10436 | null | https://arxiv.org/abs/2210.10436v2 | https://arxiv.org/pdf/2210.10436v2.pdf | LightEA: A Scalable, Robust, and Interpretable Entity Alignment Framework via Three-view Label Propagation | Entity Alignment (EA) aims to find equivalent entity pairs between KGs, which is the core step of bridging and integrating multi-source KGs. In this paper, we argue that existing GNN-based EA methods inherit the inborn defects from their neural network lineage: weak scalability and poor interpretability. Inspired by recent studies, we reinvent the Label Propagation algorithm to effectively run on KGs and propose a non-neural EA framework -- LightEA, consisting of three efficient components: (i) Random Orthogonal Label Generation, (ii) Three-view Label Propagation, and (iii) Sparse Sinkhorn Iteration. According to the extensive experiments on public datasets, LightEA has impressive scalability, robustness, and interpretability. With a mere tenth of time consumption, LightEA achieves comparable results to state-of-the-art methods across all datasets and even surpasses them on many. | ['Man Lan', 'Yuanbin Wu', 'Wenting Wang', 'Xin Mao'] | 2022-10-19 | null | null | null | null | ['entity-alignment', 'entity-alignment'] | ['knowledge-base', 'natural-language-processing'] | [ 2.41637975e-01 3.50965947e-01 -4.64598686e-01 -3.36641431e-01
-5.57211757e-01 -4.48020279e-01 1.84047118e-01 2.18896806e-01
-3.73286813e-01 8.50783646e-01 9.50330645e-02 -2.83807486e-01
-2.41563782e-01 -9.67219472e-01 -8.25726688e-01 -4.94886398e-01
-9.95798111e-02 8.72243285e-01 7.12532923e-02 -2.74455369e-01
-1.05242543e-01 -5.22533394e-02 -1.17644751e+00 4.54366505e-02
1.15004015e+00 8.98719430e-01 -3.65967721e-01 1.96355030e-01
-2.31015041e-01 1.09955823e+00 -1.04854271e-01 -1.00804055e+00
2.08844304e-01 -1.68169931e-01 -1.18169224e+00 -4.85394150e-01
3.77237290e-01 -3.78195465e-01 4.34925780e-03 1.10752761e+00
7.91240036e-01 -2.16175273e-01 5.74550867e-01 -1.65599430e+00
-8.85937333e-01 1.14648914e+00 -6.96055233e-01 -3.52624118e-01
1.17085829e-01 -1.80132031e-01 1.39170575e+00 -8.91038477e-01
9.61855352e-01 7.02885807e-01 1.39523685e+00 4.72967058e-01
-1.11027956e+00 -6.28615022e-01 2.58234382e-01 3.01394701e-01
-1.58586371e+00 -2.93902069e-01 6.07095540e-01 -1.76256131e-02
1.01773107e+00 2.25623831e-01 7.71694422e-01 1.05508578e+00
-3.93536776e-01 9.74539876e-01 8.49479735e-01 -3.90356839e-01
1.30751699e-01 -1.86171860e-01 3.28471996e-02 9.30795848e-01
5.21490037e-01 -1.83666438e-01 -6.70307994e-01 -2.48326391e-01
2.57986605e-01 -3.96918833e-01 -2.32659966e-01 -5.82049429e-01
-1.43947566e+00 7.16826975e-01 5.84496617e-01 3.99142690e-02
-4.31825131e-01 3.09247345e-01 5.81493020e-01 9.76739526e-02
5.73653400e-01 5.19436717e-01 -7.28965521e-01 5.04320078e-02
-9.69333172e-01 2.30330616e-01 9.62183774e-01 1.08541071e+00
5.86267650e-01 -2.40742788e-01 1.24160245e-01 6.30179524e-01
3.88902247e-01 3.70194525e-01 6.12464473e-02 -8.27013791e-01
5.97691238e-01 8.87011647e-01 -2.62514114e-01 -1.05431366e+00
-8.95633340e-01 -8.23403180e-01 -1.16111338e+00 -2.97301680e-01
2.71248937e-01 -2.15971366e-01 -6.82263672e-01 2.13776064e+00
6.88517034e-01 2.15687111e-01 1.71719909e-01 5.21242619e-01
1.13881695e+00 1.94044814e-01 2.60662258e-01 3.19858380e-02
1.27553642e+00 -1.52987623e+00 -5.40594697e-01 -1.74502805e-01
1.09407282e+00 -2.90214300e-01 7.41932988e-01 2.67228931e-01
-1.19787335e+00 -1.34953499e-01 -1.23239338e+00 -3.61949533e-01
-6.35893345e-01 8.15394968e-02 1.10279012e+00 5.06894410e-01
-1.28055537e+00 5.97165942e-01 -5.70380807e-01 -4.72359270e-01
4.24669147e-01 4.58007246e-01 -6.22436047e-01 -4.74909917e-02
-1.36036825e+00 7.19423532e-01 8.30763102e-01 1.72178298e-01
-2.20970199e-01 -1.09018588e+00 -8.23706269e-01 -7.12075755e-02
6.77506566e-01 -1.58147991e+00 1.01068270e+00 -4.46477234e-01
-1.18850672e+00 1.02530825e+00 7.99894929e-02 -3.12411487e-01
5.44288456e-01 -3.44279379e-01 -4.83552754e-01 3.29383351e-02
2.78588176e-01 1.07421851e+00 1.29374951e-01 -1.19175065e+00
-7.78242528e-01 -2.45541662e-01 8.64033401e-02 4.79638875e-01
-3.57264340e-01 -1.43580720e-01 -7.34579086e-01 -6.87311590e-01
4.18314844e-01 -9.44373429e-01 -1.84400812e-01 -2.34823585e-01
-9.05790389e-01 -2.85550416e-01 1.98491424e-01 -8.15868795e-01
1.49959648e+00 -1.68775856e+00 4.36547309e-01 2.78726876e-01
6.61520779e-01 1.16977297e-01 -1.60332937e-02 4.88584906e-01
-2.15475246e-01 2.24043906e-01 -2.25381568e-01 -4.58057582e-01
4.60245132e-01 2.59699253e-03 -2.26992369e-03 2.97631651e-01
-3.14038359e-02 1.36827075e+00 -1.08652341e+00 -5.61982393e-01
-3.30605149e-01 1.04929261e-01 -5.56932092e-01 -1.81649208e-01
-3.18457514e-01 2.97895044e-01 -1.27243295e-01 1.10055554e+00
8.41370285e-01 -7.68454850e-01 6.13910496e-01 -5.21122158e-01
1.56926945e-01 2.37726256e-01 -1.39998496e+00 2.11818075e+00
7.79412165e-02 8.56584087e-02 -2.16193140e-01 -7.60337830e-01
5.80195010e-01 9.73010659e-02 5.89715481e-01 -6.05341971e-01
-1.16536310e-02 5.04613698e-01 -1.80183530e-01 -2.57362843e-01
8.11426401e-01 2.23358735e-01 -2.37085775e-01 5.03082514e-01
1.00469649e-01 7.05431581e-01 4.39985514e-01 4.94690448e-01
1.09146786e+00 4.20388848e-01 4.94380713e-01 -6.68486282e-02
4.05471563e-01 -1.29072875e-01 8.75086188e-01 6.66371942e-01
4.29900996e-02 4.55649674e-01 5.09113431e-01 -7.81146526e-01
-9.29689169e-01 -8.97610307e-01 1.49520487e-01 1.22421730e+00
3.06903929e-01 -8.36479366e-01 -8.66009176e-01 -1.04616284e+00
-9.71830171e-03 5.12025297e-01 -6.55417800e-01 -1.11000769e-01
-6.73422515e-01 -1.15680981e+00 1.28420413e+00 7.79034913e-01
7.31115162e-01 -8.72682333e-01 -3.24076951e-01 2.58791953e-01
-6.93423986e-01 -1.18516195e+00 -8.89951885e-02 2.64978975e-01
-4.91933256e-01 -1.09359014e+00 -4.39254552e-01 -7.41972029e-01
7.64280379e-01 -2.21687227e-01 1.62494493e+00 2.48209670e-01
1.88334361e-01 -2.24556550e-01 -2.85374552e-01 -2.58304834e-01
-1.70500562e-01 9.17801976e-01 1.30433038e-01 -5.01716435e-01
3.72105390e-01 -8.06587398e-01 -5.88310242e-01 2.78550506e-01
-5.53050756e-01 6.07524753e-01 7.52933979e-01 6.33984447e-01
6.71952963e-01 -1.07114345e-01 7.65908301e-01 -1.29168010e+00
2.52285093e-01 -7.14247525e-01 -1.99849173e-01 7.79360652e-01
-1.13831532e+00 -9.63827223e-02 3.55742335e-01 8.47607255e-02
-8.85444164e-01 -2.42334247e-01 -3.83546859e-01 3.62208486e-02
-5.25822230e-02 7.97017276e-01 -3.22669566e-01 -1.73564196e-01
4.71570462e-01 7.12913796e-02 -3.96819949e-01 -4.62717772e-01
7.33951747e-01 3.05411965e-01 8.60979497e-01 -7.56612539e-01
7.52568722e-01 4.84192401e-01 5.15071005e-02 1.30482167e-01
-8.35533500e-01 -2.99607664e-01 -6.61417603e-01 -8.73125643e-02
7.67559052e-01 -1.16743362e+00 -7.08228230e-01 7.90931344e-01
-9.99361634e-01 -1.38188049e-01 -2.00297564e-01 1.09581515e-01
-3.67952347e-01 4.64571506e-01 -8.67534459e-01 -9.14387628e-02
-8.49689305e-01 -7.78382480e-01 9.50163901e-01 1.91187069e-01
-1.66331485e-01 -8.84536147e-01 1.67253062e-01 6.14048243e-01
4.71490026e-01 2.88917243e-01 1.17977786e+00 -9.00655150e-01
-5.78373015e-01 1.36707902e-01 -4.85881805e-01 -1.56378135e-01
-5.94040826e-02 -1.29757658e-01 -7.51439154e-01 -2.45615527e-01
-6.74302161e-01 -4.90930766e-01 8.43501151e-01 9.52313468e-02
8.36157203e-01 -2.44773075e-01 -5.72819293e-01 1.07547116e+00
1.45579135e+00 -1.01366811e-01 6.47862792e-01 8.00815225e-01
1.06382394e+00 4.34834599e-01 4.47998881e-01 2.01920256e-01
1.21421015e+00 7.06473291e-01 5.95336735e-01 -5.23954511e-01
-4.84739803e-02 -4.29200560e-01 8.14567581e-02 1.25425220e+00
-2.83007473e-01 -5.20358920e-01 -1.03344321e+00 5.58051050e-01
-2.27665281e+00 -6.27003431e-01 -2.33435512e-01 1.69599056e+00
8.39981914e-01 -1.04940124e-01 1.40273497e-01 -8.00120607e-02
6.03860140e-01 8.37149471e-02 -7.68226802e-01 -4.29962389e-02
-4.03758913e-01 -1.63259074e-01 5.44201791e-01 -2.12379456e-01
-1.21943331e+00 1.06943393e+00 6.21477079e+00 9.67044830e-01
-7.20710814e-01 1.85322970e-01 6.27778590e-01 1.32664412e-01
-4.49208230e-01 -4.63265218e-02 -9.86989737e-01 4.56214070e-01
7.18655646e-01 1.30264282e-01 3.24565679e-01 7.33148217e-01
-7.03350008e-01 1.94002777e-01 -1.04854786e+00 5.44596851e-01
6.26472011e-02 -1.49586022e+00 5.64005896e-02 6.69531599e-02
6.98907197e-01 3.76729012e-01 -1.83852673e-01 4.82292593e-01
7.96165228e-01 -9.86330628e-01 7.71153331e-01 2.59268731e-01
8.42979431e-01 -7.06058025e-01 1.09288120e+00 2.97664970e-01
-1.28092337e+00 5.27241491e-02 -1.41996533e-01 3.74881834e-01
4.99395370e-01 6.07076883e-01 -6.53985322e-01 1.41322637e+00
8.07525158e-01 6.23192668e-01 -5.79254329e-01 8.80398929e-01
-5.31705678e-01 4.00463939e-01 -5.38075805e-01 5.19455552e-01
2.30158314e-01 -1.72047898e-01 2.95642167e-01 1.12733495e+00
3.30364257e-01 -1.89259678e-01 4.87773791e-02 8.20672929e-01
-5.86947381e-01 1.83469608e-01 -4.52100933e-01 1.41406074e-01
7.37054765e-01 1.48027349e+00 -8.77285480e-01 -2.34405026e-01
-3.95472318e-01 9.57995772e-01 7.53042638e-01 1.66832313e-01
-1.19998562e+00 -2.48640522e-01 3.33706200e-01 -4.13362414e-01
3.35175395e-01 3.37032884e-01 -2.32957155e-01 -1.04815137e+00
1.20793052e-01 -8.86654556e-01 7.88683414e-01 -7.52080321e-01
-1.28518498e+00 6.40891910e-01 -1.12971947e-01 -9.10209179e-01
-2.09225386e-01 -1.48132056e-01 -3.51750702e-01 4.09745038e-01
-1.68996978e+00 -1.91314328e+00 -3.10439497e-01 3.42214644e-01
-6.32634982e-02 -5.70348352e-02 9.37797606e-01 7.93825984e-01
-8.04323196e-01 9.13815856e-01 -5.26016578e-02 2.33250827e-01
7.68767834e-01 -1.50398302e+00 9.26867604e-01 8.15326869e-01
1.19137585e-01 6.12442672e-01 2.12261960e-01 -8.40245545e-01
-1.03419995e+00 -1.22025323e+00 1.07944691e+00 -3.03829610e-01
5.52823246e-01 -3.03552836e-01 -7.00852334e-01 9.33616579e-01
3.52026731e-01 -2.90766004e-02 1.03459430e+00 5.19107461e-01
-4.91918325e-01 7.69237708e-03 -1.07963777e+00 6.21713519e-01
1.68854451e+00 -1.58744767e-01 -2.35235214e-01 1.95272863e-01
9.29460585e-01 -6.98746085e-01 -1.13715160e+00 9.89206076e-01
7.56620824e-01 -9.93818343e-01 1.21321201e+00 -6.62351966e-01
5.10399938e-01 -5.61037898e-01 -1.85061306e-01 -1.17358005e+00
-5.71357548e-01 -4.58413839e-01 -4.04139400e-01 1.63964009e+00
6.61301672e-01 -5.90691507e-01 1.03484917e+00 4.41340536e-01
-2.00492188e-01 -1.07302868e+00 -6.63404942e-01 -5.72119772e-01
-6.88372999e-02 -2.62168914e-01 1.08233011e+00 1.50812995e+00
-1.39305055e-01 4.38230097e-01 -5.48679173e-01 3.30586880e-01
8.01748037e-01 2.40237385e-01 6.82507277e-01 -1.25119460e+00
-2.51589209e-01 -2.15062603e-01 -1.24837644e-01 -7.80493796e-01
3.26426417e-01 -1.16470647e+00 -2.73009330e-01 -1.65964878e+00
5.47468960e-01 -8.44019532e-01 -5.86675823e-01 9.71153855e-01
-3.44398528e-01 5.44410646e-01 6.77869841e-02 2.98058718e-01
-1.18409431e+00 3.64580721e-01 6.62905514e-01 -6.46094009e-02
1.34518266e-01 -4.60937500e-01 -1.10652125e+00 1.04469621e+00
6.50617242e-01 -6.69107199e-01 -4.12545085e-01 -6.54128194e-01
1.25877225e+00 -3.64744008e-01 8.20028335e-02 -8.37489665e-01
6.01039946e-01 1.62165746e-01 1.86259747e-01 -8.92583787e-01
-1.05920181e-01 -5.54206610e-01 7.57644653e-01 3.66097659e-01
-2.53660679e-01 3.61940712e-01 -1.69107229e-01 4.47231919e-01
-3.86185870e-02 -1.32393464e-01 2.19555750e-01 -1.21349066e-01
-1.00871134e+00 2.87918210e-01 2.00577855e-01 2.39357725e-01
9.36413467e-01 -2.94119298e-01 -9.68991220e-01 4.56335805e-02
-5.13927698e-01 5.07618427e-01 5.56113958e-01 1.09956980e-01
1.94795251e-01 -1.50132835e+00 -7.85329044e-01 5.97629398e-02
1.33555278e-01 4.01635945e-01 3.00908178e-01 1.17656350e+00
-5.66031337e-01 2.44674101e-01 -1.03455089e-01 -3.56443018e-01
-1.19608212e+00 1.87048391e-01 9.86549929e-02 -9.51315284e-01
-6.02081835e-01 1.14402223e+00 3.68120102e-03 -1.13996506e+00
1.62688211e-01 8.40556920e-02 -1.92660123e-01 1.76365361e-01
8.87023062e-02 5.16073644e-01 2.49585822e-01 -5.19086421e-01
-5.07156014e-01 5.15604436e-01 -4.52258825e-01 4.24137086e-01
1.44483995e+00 -8.33816528e-02 -5.56648195e-01 -2.50597876e-02
8.57270598e-01 1.49360582e-01 -6.38278902e-01 -2.78523982e-01
2.46772081e-01 1.28878891e-01 -2.53698677e-01 -1.21067965e+00
-1.25557446e+00 2.55978882e-01 2.36228406e-01 -4.99121659e-02
1.30488348e+00 1.86474167e-03 1.22135317e+00 4.98109907e-01
5.29254496e-01 -1.11389828e+00 -3.12664568e-01 3.56218219e-01
2.49266565e-01 -1.06267941e+00 2.08364949e-01 -8.89905691e-01
-4.20764893e-01 6.88881814e-01 7.50039577e-01 4.64644879e-01
1.74649492e-01 2.66084105e-01 -1.61544517e-01 -5.29085755e-01
-8.28194499e-01 -1.42854795e-01 1.04358062e-01 5.12644291e-01
2.47421721e-03 -1.45077392e-01 -1.66550040e-01 8.51104379e-01
-1.40169531e-01 1.96174100e-01 3.59131917e-02 8.03666651e-01
1.06239982e-01 -1.23657775e+00 2.80282795e-01 2.59030014e-01
-5.41265428e-01 -5.05983353e-01 -4.41444010e-01 1.01492274e+00
6.03985369e-01 5.09151280e-01 -2.56821305e-01 -5.85269034e-01
2.73366690e-01 1.33539304e-01 2.56456673e-01 -3.30539867e-02
-8.83977294e-01 -4.93452311e-01 5.14976323e-01 -6.05607927e-01
-6.73787057e-01 -5.32689691e-01 -1.28487182e+00 -6.40941620e-01
-6.02912247e-01 1.39807258e-02 6.06098890e-01 9.32305396e-01
7.84683406e-01 4.45822924e-01 2.26686686e-01 -4.06609505e-01
-2.70850182e-01 -7.67354190e-01 -3.69618714e-01 3.67051899e-01
-1.96365640e-01 -4.98674810e-01 7.99510255e-03 -2.46473372e-01] | [8.800410270690918, 8.035035133361816] |
37e3995b-b8cc-436f-962e-13091f92c073 | validation-loss-for-landmark-detection | 1901.10143 | null | https://arxiv.org/abs/1901.10143v3 | https://arxiv.org/pdf/1901.10143v3.pdf | Learning to Validate the Quality of Detected Landmarks | We present a new loss function for the validation of image landmarks detected via Convolutional Neural Networks (CNN). The network learns to estimate how accurate its landmark estimation is. This loss function is applicable to all regression-based location estimations and allows the exclusion of unreliable landmarks from further processing. In addition, we formulate a novel batch balancing approach which weights the importance of samples based on their produced loss. This is done by computing a probability distribution mapping on an interval from which samples can be selected using a uniform random selection scheme. We conducted experiments on the 300W, AFLW, and WFLW facial landmark datasets. In the first experiments, the influence of our batch balancing approach is evaluated by comparing it against uniform sampling. In addition, we evaluated the impact of the validation loss on the landmark accuracy based on uniform sampling. The last experiments evaluate the correlation of the validation signal with the landmark accuracy. All experiments were performed for all three datasets. | ['Wolfgang Fuhl', 'Enkelejda Kasneci'] | 2019-01-29 | null | null | null | null | ['head-pose-estimation'] | ['computer-vision'] | [-4.85843569e-02 1.17132455e-01 -2.40113363e-01 -8.04965556e-01
-1.00743747e+00 -2.49838412e-01 6.53739572e-01 3.77481490e-01
-8.85308683e-01 6.56392574e-01 -1.19143896e-01 8.03767741e-02
-2.01181471e-01 -6.80241644e-01 -8.45310271e-01 -7.17760026e-01
-3.75923932e-01 1.98987544e-01 1.66122571e-01 3.29751432e-01
2.36568898e-01 8.15076828e-01 -1.24857640e+00 -1.31642595e-01
4.63873237e-01 1.39613771e+00 -3.58835876e-01 2.09725276e-01
2.31486395e-01 4.05846030e-01 -7.81799138e-01 -2.98078388e-01
4.25820917e-01 -2.80914903e-01 -3.72963279e-01 -2.11134657e-01
6.72618985e-01 -1.86507449e-01 2.46977717e-01 9.18713510e-01
7.53451407e-01 1.87402099e-01 7.70765603e-01 -1.29540741e+00
-4.97271791e-02 4.27531749e-01 -6.62169039e-01 8.36360604e-02
1.89090610e-01 -2.64403641e-01 9.12316084e-01 -9.56470132e-01
5.52883446e-01 9.73130226e-01 1.02642167e+00 3.07652831e-01
-1.38759768e+00 -8.35971534e-01 -7.33843818e-02 1.11675642e-01
-1.77171314e+00 -6.48891270e-01 7.75009453e-01 -3.58531743e-01
1.69215247e-01 -1.79966569e-01 1.96280435e-01 4.81554866e-01
1.24137498e-01 4.39111769e-01 7.33853638e-01 -7.29633451e-01
4.44082320e-01 1.12171054e-01 -1.33495584e-01 7.27310896e-01
2.50830591e-01 3.02350491e-01 -6.28519773e-01 -1.83795601e-01
6.14617109e-01 -5.04544973e-01 -2.35613212e-01 -3.92153859e-01
-6.84866190e-01 8.02012205e-01 6.81527555e-01 2.67286569e-01
-2.22763911e-01 3.22357088e-01 3.57051075e-01 6.59875274e-02
7.02204227e-01 1.87886164e-01 -2.84595072e-01 2.77203411e-01
-1.44056559e+00 -2.56619453e-01 4.73706543e-01 5.39283752e-01
8.06553423e-01 -7.88921788e-02 -3.81293982e-01 8.06000948e-01
3.61499578e-01 3.07369202e-01 3.56110275e-01 -7.39356577e-01
5.87933930e-04 2.65544176e-01 1.00916311e-01 -1.03177905e+00
-6.50249302e-01 -3.81216645e-01 -5.39257884e-01 6.10753775e-01
6.89617753e-01 -2.78356582e-01 -9.00320351e-01 1.79037094e+00
2.56356895e-01 2.94481456e-01 -4.24092710e-01 8.42923105e-01
6.64346159e-01 2.18114436e-01 1.37159124e-01 4.13905829e-02
8.67902637e-01 -3.43063205e-01 -2.97102273e-01 4.14280266e-01
2.54090965e-01 -7.51990080e-01 8.27623367e-01 2.27566883e-01
-8.15397739e-01 -6.94313705e-01 -1.18709457e+00 1.51363641e-01
-1.35978118e-01 7.31698632e-01 1.91030502e-01 7.44615436e-01
-1.10278380e+00 8.40489745e-01 -8.97285223e-01 -1.29102349e-01
5.14352083e-01 6.06861651e-01 -3.55239719e-01 3.66655856e-01
-9.02237415e-01 8.42749238e-01 -3.85857038e-02 3.63658458e-01
-6.31078839e-01 -7.47432828e-01 -8.42909992e-01 8.93798098e-02
-1.88664302e-01 -2.39762873e-03 8.16877604e-01 -1.16119432e+00
-1.43569386e+00 8.71527016e-01 -1.89513117e-01 -6.84911728e-01
8.84690940e-01 2.65176222e-02 -1.18442103e-01 6.88659539e-03
1.15897588e-01 6.74959421e-01 9.04342592e-01 -1.09604347e+00
-4.79831666e-01 -3.80978465e-01 -2.05282673e-01 -3.55353624e-01
2.64971498e-02 -2.98595876e-01 -4.14651781e-01 -4.50163722e-01
1.78525344e-01 -7.63383627e-01 -3.59807424e-02 3.37016493e-01
-4.14137870e-01 -1.15726613e-01 3.21133047e-01 -3.60031813e-01
8.65621150e-01 -2.37059236e+00 -4.79102045e-01 9.28531528e-01
-8.46949816e-02 -9.56631675e-02 -3.26757371e-01 9.60552618e-02
-3.19448486e-02 1.06554374e-01 -3.85677442e-03 -4.85723644e-01
-1.64213449e-01 -2.64685571e-01 -1.69466957e-01 9.14291561e-01
2.82029390e-01 3.32825452e-01 -5.55217564e-01 -4.99282926e-01
8.68812352e-02 8.01021218e-01 -4.10345256e-01 -9.10945311e-02
1.29239902e-01 3.41478318e-01 -2.48828992e-01 5.22978127e-01
6.78976417e-01 9.90291759e-02 5.52866701e-03 -3.97551358e-01
-2.07191072e-02 -8.91046599e-02 -1.09072089e+00 1.29842055e+00
-8.05443525e-01 8.42860341e-01 -1.04478694e-01 -7.06505477e-01
1.38641262e+00 1.87899381e-01 5.45670450e-01 -6.84047818e-01
2.16449678e-01 2.69286305e-01 -9.53242183e-02 1.13570005e-01
1.89967349e-01 -5.11184074e-02 1.75762475e-01 2.49411777e-01
1.92752138e-01 1.28516361e-01 1.25160947e-01 -3.30607027e-01
6.73495889e-01 1.64969295e-01 1.31996855e-01 -4.84121978e-01
6.55738831e-01 -4.41429734e-01 5.64968944e-01 6.27342522e-01
-2.98238039e-01 7.53490269e-01 7.66866088e-01 -5.44670045e-01
-6.81547940e-01 -1.03214324e+00 -6.46541476e-01 6.65646255e-01
1.62275419e-01 -1.28842488e-01 -5.15323937e-01 -7.55829990e-01
1.71137318e-01 5.25445104e-01 -9.35188711e-01 -4.05836046e-01
-3.21560234e-01 -7.57057428e-01 6.91271961e-01 4.44241494e-01
4.95078355e-01 -8.35426331e-01 -6.11426353e-01 -1.62161887e-01
1.41359344e-01 -8.09923589e-01 -4.12558615e-01 2.16378748e-01
-6.01281524e-01 -1.25035286e+00 -7.04497337e-01 -6.36349857e-01
9.86323297e-01 -4.08324510e-01 9.60403919e-01 1.08360849e-01
-3.04837465e-01 3.29052359e-01 -1.41433537e-01 -4.07133192e-01
-8.50453004e-02 3.17664653e-01 -8.58928487e-02 3.34314436e-01
3.55136991e-01 -3.23160529e-01 -5.78236520e-01 5.04611194e-01
-4.65575218e-01 -7.58947134e-01 2.13680074e-01 8.82569075e-01
7.66886652e-01 5.25102764e-02 6.09133124e-01 -5.88493168e-01
5.87430477e-01 -1.60237506e-01 -1.04296231e+00 2.49735355e-01
-4.38757002e-01 3.26336056e-01 4.08990085e-01 -4.19104397e-01
-6.97153270e-01 2.05130324e-01 -2.01073095e-01 -1.55516058e-01
4.90724854e-02 3.70208532e-01 1.84206307e-01 -5.41416347e-01
6.86277390e-01 -2.59361506e-01 8.90322253e-02 -2.57778853e-01
1.41448990e-01 2.81629294e-01 2.74796069e-01 -5.98918378e-01
5.76050997e-01 5.00138462e-01 4.37820882e-01 -9.51838195e-01
-5.47947705e-01 -2.66297549e-01 -5.91065228e-01 -4.66654241e-01
4.98099625e-01 -4.35979903e-01 -8.90332758e-01 2.47651547e-01
-8.92390192e-01 -3.86605769e-01 -5.35566986e-01 6.71560168e-01
-4.12916780e-01 9.28913131e-02 -1.18074715e-01 -8.91622901e-01
-1.39123172e-01 -1.11300373e+00 1.08411849e+00 2.48398677e-01
-4.26888227e-01 -8.22774947e-01 7.74126127e-02 -3.22746634e-01
3.69909078e-01 5.02236366e-01 6.31236374e-01 -6.05144262e-01
-1.92919314e-01 -6.28530860e-01 -2.59621561e-01 3.81090581e-01
-4.76327054e-02 3.66484642e-01 -1.12695670e+00 -2.46604726e-01
-2.59773791e-01 -1.70673132e-01 1.04025030e+00 7.71696568e-01
9.82143998e-01 2.17338681e-01 -1.56837329e-01 5.56023121e-01
1.52682853e+00 1.53355315e-01 6.27956688e-01 3.55512351e-01
5.48670404e-02 4.04961079e-01 5.39459765e-01 3.60362560e-01
-1.84686527e-01 7.92512417e-01 4.01854008e-01 -1.88174516e-01
-1.37047723e-01 -1.22527257e-01 3.68758105e-02 -7.98381418e-02
-1.35018051e-01 -4.69735488e-02 -7.74856389e-01 4.66066539e-01
-1.32635295e+00 -6.59131169e-01 2.99450427e-01 2.69458985e+00
6.50839090e-01 3.15939069e-01 1.62040442e-01 2.30376601e-01
7.43247569e-01 3.99982296e-02 -2.44223744e-01 -3.45507264e-01
2.56169975e-01 4.73032594e-01 7.27685094e-01 6.86358988e-01
-1.18033659e+00 6.18110240e-01 6.49504185e+00 8.47798467e-01
-1.56624901e+00 -1.52371436e-01 1.06121790e+00 -2.16286227e-01
1.54253766e-01 -3.68378252e-01 -8.87855113e-01 5.35752237e-01
6.32676601e-01 9.53690708e-02 -1.66254371e-01 8.01483214e-01
2.39846796e-01 -5.12825966e-01 -8.72093916e-01 6.52889907e-01
2.03421395e-02 -9.73825216e-01 -2.64009386e-01 -1.78852528e-01
4.64711636e-01 9.86597538e-02 1.15867697e-01 -7.17369765e-02
-1.15244977e-01 -9.33907688e-01 8.16525578e-01 4.88511175e-01
9.66024399e-01 -9.25343454e-01 1.00207758e+00 -5.30906990e-02
-9.90301907e-01 1.18042856e-01 -3.28138351e-01 3.93862098e-01
-6.62630200e-02 9.27323043e-01 -9.85280275e-01 2.74018139e-01
5.02608895e-01 1.09692708e-01 -5.53887010e-01 1.35318470e+00
-5.42041421e-01 4.97753918e-01 -6.38230026e-01 -1.84131786e-01
2.92421784e-02 -1.02221057e-01 2.12326169e-01 1.06771493e+00
2.75120437e-01 -5.46086729e-01 -2.69410964e-02 7.59620428e-01
-1.80160180e-01 1.62413895e-01 -2.72978604e-01 5.02697587e-01
5.18455207e-01 1.21570003e+00 -9.95052993e-01 2.16009036e-01
-7.45880827e-02 5.85611105e-01 3.38255227e-01 2.99518794e-01
-7.86380589e-01 -5.68291008e-01 3.63084733e-01 3.67431670e-01
2.43538544e-01 -1.46537706e-01 -2.19348922e-01 -5.04728615e-01
9.79060754e-02 -5.43049164e-02 2.01808602e-01 -3.33940327e-01
-1.11464608e+00 7.68992066e-01 3.01695969e-02 -1.06567061e+00
-1.82675377e-01 -4.25763220e-01 -7.49168634e-01 7.82362819e-01
-1.57887673e+00 -7.19538450e-01 -2.63487905e-01 1.87767595e-01
-1.82179995e-02 4.14154405e-04 6.13900363e-01 5.16534209e-01
-4.61108506e-01 1.00522459e+00 -8.03529173e-02 4.70007777e-01
7.71584988e-01 -9.22214210e-01 9.15043280e-02 6.57598019e-01
2.07102939e-01 5.00274599e-01 4.93573636e-01 -2.63823092e-01
-4.02882278e-01 -1.03778350e+00 8.60210538e-01 -5.99328391e-02
1.15750834e-01 -1.82017982e-01 -5.47671974e-01 5.13374865e-01
-3.30998212e-01 4.00906444e-01 4.91555333e-01 2.13763401e-01
-3.07629496e-01 -4.93340999e-01 -1.54740322e+00 2.16426209e-01
4.78916049e-01 -2.95956045e-01 7.97728822e-02 -5.17488569e-02
6.40572086e-02 -2.36698672e-01 -7.84199774e-01 5.49093544e-01
9.39272404e-01 -1.01501226e+00 9.22953069e-01 -2.03654125e-01
1.60993680e-01 -3.30472678e-01 3.67846601e-02 -1.21760166e+00
-8.88644084e-02 -1.70028552e-01 5.58722496e-01 1.23665953e+00
6.25677884e-01 -5.62950075e-01 9.38288391e-01 4.58501041e-01
3.82125199e-01 -9.39787447e-01 -1.38446009e+00 -8.23610187e-01
-1.64277345e-01 -3.35341841e-01 5.69229424e-01 4.73275989e-01
-3.30910176e-01 -2.35656753e-01 -1.81169063e-02 2.74328887e-02
6.36110902e-01 -5.68306930e-02 5.70280492e-01 -1.17989850e+00
1.29028708e-01 -4.79808837e-01 -8.86055768e-01 -4.25897777e-01
2.47623622e-01 -7.43758500e-01 2.66591698e-01 -9.99273181e-01
-3.28041613e-01 -5.60326099e-01 -5.03346324e-01 4.42488849e-01
2.77621657e-01 7.30769515e-01 6.85877353e-02 -9.16778948e-03
-2.76719868e-01 3.71766031e-01 7.51410365e-01 -1.18924463e-02
-2.77506709e-01 3.67873788e-01 -9.77371782e-02 7.90539861e-01
8.37391973e-01 -4.56433982e-01 -2.11683333e-01 -1.18700191e-01
9.47871730e-02 -2.88634926e-01 4.27681237e-01 -1.19998527e+00
1.60915688e-01 3.72666061e-01 7.61156261e-01 -5.17631114e-01
9.63091329e-02 -9.50887442e-01 -1.41995385e-01 2.33823150e-01
-5.89521170e-01 -3.09383392e-01 3.25633109e-01 2.38434836e-01
-2.56201923e-01 -3.81276667e-01 1.21630156e+00 3.19507003e-01
-4.78932619e-01 9.52662453e-02 7.27931187e-02 -5.93904592e-02
9.27989185e-01 -1.89584851e-01 2.59230196e-01 -4.94648904e-01
-7.22755969e-01 6.38616309e-02 2.10794717e-01 1.08362682e-01
4.49750751e-01 -1.29375708e+00 -5.79494119e-01 5.95508873e-01
1.55884519e-01 -2.89095521e-01 -3.67130071e-01 7.94413209e-01
-5.46628833e-01 6.28233328e-02 -1.82557821e-01 -5.08291304e-01
-1.15296793e+00 1.45703018e-01 6.28049195e-01 -4.77394499e-02
-1.71967838e-02 8.91373098e-01 -3.24822754e-01 -2.70163596e-01
6.36652589e-01 -4.87220377e-01 -1.56547993e-01 1.77536592e-01
4.34998274e-01 5.81638038e-01 3.93458486e-01 -5.85145831e-01
-5.39680541e-01 7.26095974e-01 2.24605039e-01 -1.74245179e-01
1.16682327e+00 1.24066547e-01 -1.10270521e-02 3.20684195e-01
1.37416923e+00 1.92156866e-01 -1.16029143e+00 2.25845464e-02
1.43044695e-01 -3.21284711e-01 1.62946507e-01 -6.36781573e-01
-1.34108078e+00 6.42539799e-01 8.66159618e-01 -2.66490225e-02
1.08180594e+00 -2.96153307e-01 2.34630361e-01 -8.53429884e-02
3.67228687e-01 -1.06849611e+00 -3.14553350e-01 1.80660412e-01
7.69623935e-01 -1.24220407e+00 2.46049583e-01 -1.84681386e-01
-2.38646064e-02 1.36650097e+00 3.50316733e-01 -3.74835432e-01
1.08803844e+00 3.24740529e-01 1.85040966e-01 -7.23258331e-02
-6.17821962e-02 -3.15577835e-02 4.60651755e-01 4.81079191e-01
6.17137313e-01 -1.05671093e-01 -5.64897418e-01 3.02013010e-01
-8.04941207e-02 3.54504555e-01 1.05885729e-01 4.94308949e-01
-9.60803404e-02 -9.54076827e-01 -2.61718601e-01 3.19290370e-01
-4.98615175e-01 2.09548965e-01 -2.41498038e-01 1.02410781e+00
3.24428111e-01 7.64816463e-01 3.31723779e-01 -2.11609647e-01
2.64415771e-01 -7.29172528e-02 4.47421491e-01 -1.66914627e-01
-4.09339994e-01 -2.85563171e-02 -1.20739728e-01 -5.83625734e-01
-3.94645363e-01 -5.21418512e-01 -1.20320237e+00 -2.88018975e-02
-6.64100587e-01 2.88557887e-01 9.17653203e-01 5.36085963e-01
7.69413710e-02 2.26339474e-01 6.68951273e-01 -8.90911877e-01
-7.00631738e-01 -8.09754729e-01 -7.10940063e-01 1.76503256e-01
4.33287680e-01 -7.92388678e-01 -5.91276526e-01 -2.59450972e-01] | [13.34043025970459, 0.6962951421737671] |
98be75d4-baa3-4bc5-bebd-f6da8fafb5fc | an-efficientnet-based-modified-sigmoid | null | null | https://doi.org/10.1016/j.cmpb.2022.106935 | https://www.sciencedirect.com/science/article/pii/S0169260722003170?via%3Dihub | An EfficientNet-based modified sigmoid transform for enhancing dermatological macro-images of melanoma and nevi skin lesions | Background and objective: During the initial stages, skin lesions may not have sufficient intensity difference or contrast from the background region on dermatological macro-images. The lack of proper light exposure at the time of capturing the image also reduces the contrast. Low contrast between lesion and background regions adversely impacts segmentation. Enhancement techniques for improving the contrast between lesion and background skin on dermatological macro-images are limited in the literature. An EfficientNet-based modified sigmoid transform for enhancing the contrast on dermatological macroimages is proposed to address this issue.
Methods: A modified sigmoid transform is applied in the HSV color space. The crossover point in the modified sigmoid transform that divides the macro-image into lesion and background is predicted using a modified EfficientNet regressor to exclude manual intervention and subjectivity. The Modified EfficientNet regressor is constructed by replacing the classifier layer in the conventional EfficientNet with a regression layer. Transfer learning is employed to reduce the training time and size of the dataset required to train
the modified EfficientNet regressor. For training the modified EfficientNet regressor, a set of value components extracted from the HSV color space representation of the macro-images in the training dataset is fed as input. The corresponding set of ideal crossover points at which the values of Dice similarity coefficient (DSC) between the ground-truth images and the segmented output images obtained from Otsu’s thresholding are maximum is defined as the target.
Results: On images enhanced with the proposed framework, the DSC of segmented results obtained by Otsu’s thresholding increased from 0.68 ± 0.34 to 0.81 ± 0.17.
Conclusions: The proposed algorithm could consistently improve the contrast between lesion and background on a comprehensive set of test images, justifying its applications in automated analysis of dermatological macro-images. | ['Malaya Kumar Nath', 'M. Vipin Das', 'Justin Joseph', 'Vipin Venugopal'] | 2022-07-17 | null | null | null | computer-methods-and-programs-in-biomedicine-3 | ['local-color-enhancement', 'skin-lesion-segmentation'] | ['computer-vision', 'medical'] | [ 6.52666211e-01 -6.53971434e-02 -5.59905432e-02 -2.60220379e-01
-1.57904223e-01 -3.06012124e-01 1.58657283e-01 8.07042494e-02
-7.63319850e-01 7.87670672e-01 -7.26069272e-01 -2.48180285e-01
-6.26023039e-02 -7.77249157e-01 -1.46510214e-01 -1.13027453e+00
3.21465760e-01 -1.33102834e-01 3.94064426e-01 1.56758949e-02
4.05175358e-01 7.18079448e-01 -1.28981483e+00 4.16186154e-01
1.24686193e+00 1.05037749e+00 3.78778905e-01 8.61061871e-01
-1.76848739e-01 3.75335604e-01 -8.09537768e-01 -3.50826234e-02
3.71567190e-01 -8.21536541e-01 -4.27982867e-01 3.08455110e-01
3.44904333e-01 -2.75036454e-01 -2.28003040e-02 1.18139601e+00
5.47189713e-01 1.72556221e-01 8.60095322e-01 -1.04136062e+00
-4.97555912e-01 -2.17538893e-01 -8.89550149e-01 4.48681206e-01
-1.04392275e-01 2.30146885e-01 2.21701652e-01 -5.43288648e-01
6.77963078e-01 8.44212949e-01 4.68165040e-01 4.24020797e-01
-1.21996284e+00 -5.05610228e-01 -2.65335351e-01 4.38202590e-01
-1.51528060e+00 2.31674567e-01 4.63662624e-01 -3.66623014e-01
7.62547731e-01 7.00117588e-01 1.03013468e+00 4.28737044e-01
7.27181375e-01 2.30040818e-01 1.76094329e+00 -6.18791401e-01
1.29813151e-02 6.73200727e-01 -1.42598376e-01 9.56969798e-01
2.52203792e-01 2.64190555e-01 9.08630714e-02 2.43422538e-01
1.08859289e+00 -3.67868952e-02 -5.31846210e-02 1.35282815e-01
-4.55545098e-01 5.98896563e-01 5.50961196e-01 6.05658650e-01
-4.59344178e-01 -3.78502816e-01 1.31518960e-01 1.50130495e-01
2.95641541e-01 2.34044418e-01 -2.25915035e-04 3.68030429e-01
-1.02060640e+00 -3.93729001e-01 5.11167824e-01 2.86770791e-01
5.64355433e-01 1.20122910e-01 -7.67503604e-02 9.03042614e-01
-5.90177886e-02 5.04727066e-01 4.24729437e-01 -6.90377116e-01
-8.69900882e-02 1.04052138e+00 -2.57409155e-01 -1.06023192e+00
-1.61632136e-01 -3.36005777e-01 -7.94189632e-01 7.44458556e-01
6.49946451e-01 -1.71185866e-01 -1.29411268e+00 9.85978246e-01
6.88920975e-01 1.42819295e-02 -1.16953239e-01 1.02675903e+00
6.64091825e-01 6.58122122e-01 1.69148102e-01 -3.63213718e-01
1.22338390e+00 -7.06330538e-01 -6.33474469e-01 9.12532359e-02
3.45609426e-01 -9.65878606e-01 1.13184464e+00 4.01515365e-01
-1.14860308e+00 -5.58110535e-01 -1.15303719e+00 2.92927921e-01
-4.11262274e-01 4.07609284e-01 1.73149869e-01 7.21021652e-01
-9.41646934e-01 5.97065330e-01 -7.59467721e-01 -4.31034714e-01
1.96612552e-01 4.83651668e-01 -3.84236813e-01 1.06958766e-02
-8.16362977e-01 1.28051996e+00 4.77681309e-01 4.96298313e-01
-2.47189865e-01 -4.58890229e-01 -5.67493081e-01 -1.59571469e-01
-1.07111118e-03 -3.42990428e-01 5.18669009e-01 -1.58167672e+00
-1.34597909e+00 1.11230302e+00 1.29982710e-01 -1.51649639e-01
5.21366537e-01 5.43569684e-01 -4.19851452e-01 5.84880412e-01
-2.71586448e-01 4.62272048e-01 8.32391739e-01 -1.32193375e+00
-9.38607156e-01 -2.80707866e-01 -1.88298166e-01 4.02418286e-01
-1.08540252e-01 2.87786648e-02 -3.69955450e-01 -4.40613627e-01
-1.71242002e-02 -8.53801489e-01 -2.52239645e-01 2.13207290e-01
-3.65647286e-01 1.81664690e-01 9.38071132e-01 -1.18089080e+00
1.07028806e+00 -2.06104708e+00 -4.91969347e-01 7.79087842e-01
1.87254369e-01 5.94863296e-01 -7.61361942e-02 -2.38597080e-01
-1.77129179e-01 -2.35437639e-02 -4.19553280e-01 5.74023545e-01
-5.04996896e-01 6.93448484e-02 6.28570735e-01 7.02032626e-01
7.43182451e-02 5.45286238e-01 -6.29390478e-01 -9.91158128e-01
5.98230898e-01 6.82559788e-01 9.68791731e-03 7.82674700e-02
2.35488161e-01 3.05111021e-01 -4.15595174e-02 7.40141630e-01
9.81471717e-01 -4.01545316e-02 1.49477214e-01 -3.54803562e-01
-5.69825470e-02 -4.32543993e-01 -1.14740324e+00 8.17171693e-01
-3.57417762e-01 7.99802363e-01 1.40665203e-01 -8.74315381e-01
1.01754749e+00 3.00610691e-01 6.59006476e-01 -1.03226864e+00
4.68579859e-01 1.80816352e-01 3.94214094e-01 -8.54724765e-01
1.02245726e-01 -3.71802181e-01 5.40705383e-01 2.67885894e-01
-1.27984807e-01 -2.65060097e-01 4.25453871e-01 -2.89846331e-01
4.97200608e-01 -4.76865135e-02 4.10032094e-01 -2.42352337e-02
8.85188520e-01 1.61371022e-01 3.11533332e-01 5.20227432e-01
-3.88771981e-01 3.36391270e-01 3.76121342e-01 -2.02091843e-01
-1.01143098e+00 -1.04374480e+00 -3.49447101e-01 6.36967599e-01
2.73282081e-01 5.78774393e-01 -1.13095427e+00 -6.55876398e-01
-2.22791791e-01 4.86536205e-01 -6.53097928e-01 -1.31389394e-01
-4.08360034e-01 -9.04931724e-01 4.52624261e-01 2.10294053e-01
7.04347312e-01 -1.16897511e+00 -8.18828225e-01 -2.98028961e-02
-9.64633003e-02 -6.35682762e-01 -3.02619696e-01 -2.38727476e-03
-9.43210304e-01 -1.41351366e+00 -7.38987446e-01 -1.00225163e+00
1.22274315e+00 4.76799123e-02 4.76583034e-01 3.09039176e-01
-9.98217762e-01 -6.74592517e-03 3.71481515e-02 -3.97888094e-01
-5.65441489e-01 -3.87626886e-01 -5.07874310e-01 -1.63615588e-02
3.61281633e-01 -7.71016777e-02 -8.21928918e-01 2.14772716e-01
-1.02242267e+00 -2.87725274e-02 6.99907482e-01 8.36332440e-01
6.43212974e-01 2.21000016e-01 1.73054144e-01 -6.43583775e-01
7.37188637e-01 9.86289978e-02 -7.06609368e-01 4.85208184e-01
-6.95425034e-01 -2.59789348e-01 6.16621256e-01 -6.10228300e-01
-1.24494481e+00 -2.31710792e-01 2.04578027e-01 -1.20339967e-01
-1.84777185e-01 2.45206311e-01 4.67551053e-01 -4.76796329e-01
8.97893190e-01 1.25390097e-01 3.65599155e-01 1.12198010e-01
-5.78277074e-02 8.24371994e-01 5.77467799e-01 9.01628733e-02
4.78650063e-01 4.82173026e-01 1.59318238e-01 -8.72246683e-01
-2.80261636e-01 -3.88633430e-01 -5.21225274e-01 -6.39817894e-01
9.55496907e-01 -1.16022654e-01 -6.37596190e-01 5.26359260e-01
-8.54534626e-01 -2.63826430e-01 -2.15243369e-01 6.39034331e-01
-1.48071602e-01 3.56172383e-01 -6.78760827e-01 -9.14169788e-01
-5.59163988e-01 -1.15002608e+00 2.56165087e-01 8.34167004e-01
5.71099184e-02 -1.38300860e+00 -1.73389316e-01 2.36453652e-01
2.34031677e-01 6.44128919e-01 1.09312713e+00 -2.98421741e-01
-1.60580754e-01 -5.03094375e-01 -3.37751836e-01 7.80700088e-01
4.55567718e-01 4.72026587e-01 -8.49630356e-01 -6.91381916e-02
2.25596562e-01 1.71989888e-01 6.85360432e-01 9.06931281e-01
1.04936981e+00 -1.77062824e-01 -5.61992824e-02 6.77566111e-01
1.88340974e+00 8.97264898e-01 7.47041285e-01 4.07831877e-01
3.00114006e-01 8.45546484e-01 8.70763779e-01 3.15765925e-02
-2.27883279e-01 1.63376227e-01 2.05698162e-01 -8.82201731e-01
-3.12624127e-01 1.98739558e-01 2.06663143e-02 3.63109142e-01
-4.46493745e-01 6.20559976e-02 -5.70361435e-01 4.50490922e-01
-1.06921852e+00 -8.18164527e-01 -2.63391435e-01 2.22716117e+00
9.01542842e-01 4.38896604e-02 7.62772858e-02 2.18582571e-01
1.23072779e+00 -4.22333091e-01 -5.85825384e-01 -9.70723391e-01
-1.65345103e-01 6.65484190e-01 6.18599832e-01 6.51348233e-01
-8.93173039e-01 6.63465083e-01 5.74632311e+00 8.41127515e-01
-1.79019749e+00 -2.50070423e-01 7.13446081e-01 -4.19059657e-02
1.17168762e-01 -2.38863781e-01 -3.01017731e-01 5.32407463e-01
6.28254235e-01 -1.18162312e-01 4.67268378e-01 4.56766248e-01
5.50692737e-01 -7.94631720e-01 -4.95006353e-01 8.86150479e-01
4.53271642e-02 -9.55987453e-01 -1.31373674e-01 2.08225176e-02
8.58645618e-01 -6.11063004e-01 4.07377005e-01 -2.83114403e-01
-2.19416603e-01 -1.14477980e+00 7.58119747e-02 6.28164947e-01
1.11896992e+00 -7.71605968e-01 1.11349380e+00 -3.71811837e-02
-7.68095076e-01 2.80969422e-02 -2.25994065e-01 3.69589657e-01
-2.11329341e-01 4.18279320e-01 -1.48536265e+00 1.92904532e-01
3.32241744e-01 3.38323042e-03 -6.40495777e-01 1.11703074e+00
-1.11207902e-01 4.33763951e-01 -2.28409290e-01 -2.09117800e-01
4.27807480e-01 -6.66115046e-01 4.24716830e-01 1.30151856e+00
1.90020338e-01 -6.48440942e-02 -3.69016618e-01 9.25005257e-01
4.25310105e-01 4.64478463e-01 -3.95656168e-01 5.18085025e-02
2.14032903e-01 1.46149004e+00 -1.14744115e+00 -3.56638700e-01
-2.01491624e-01 9.14301634e-01 -4.62800920e-01 5.95641375e-01
-8.05998325e-01 -6.36278331e-01 -9.74015221e-02 2.66801298e-01
-4.45755385e-02 3.37485552e-01 -8.42609942e-01 -4.33295459e-01
-7.10843876e-02 -7.29066372e-01 6.76491022e-01 -7.31352568e-01
-9.11368668e-01 6.85274243e-01 -8.25423077e-02 -1.12589896e+00
-9.48862433e-02 -6.79782391e-01 -7.07626820e-01 1.21168530e+00
-1.23612201e+00 -9.45952892e-01 -6.86794758e-01 6.84514046e-01
2.26877153e-01 -9.45933610e-02 6.36519313e-01 -2.82444004e-02
-5.06695986e-01 5.40108919e-01 9.22880098e-02 -1.49384752e-01
7.31146574e-01 -1.33736730e+00 -4.67480719e-01 8.86230648e-01
-5.53433478e-01 5.06689787e-01 6.38541698e-01 -7.31503487e-01
-4.33822662e-01 -7.16194391e-01 5.64195335e-01 2.80906439e-01
2.60734469e-01 1.64574459e-01 -7.43508577e-01 1.22640252e-01
2.87516475e-01 -1.44619077e-01 7.61714697e-01 -6.84763014e-01
2.31314003e-01 -1.79840177e-01 -1.90402663e+00 7.92769730e-01
-6.39279280e-03 -2.78996944e-01 -2.30128825e-01 1.30964845e-01
-1.01723716e-01 -3.87616575e-01 -9.61448014e-01 1.60646886e-01
8.60683799e-01 -8.13483298e-01 8.42383683e-01 -3.34105402e-01
6.07636094e-01 -1.79989651e-01 2.60309935e-01 -1.08171952e+00
6.32559136e-03 -1.49098128e-01 3.28354836e-01 6.07445300e-01
3.72478813e-01 -6.56298041e-01 9.12363768e-01 4.21638995e-01
2.44306520e-01 -1.07832611e+00 -7.26651549e-01 -4.10670936e-01
-1.41725078e-01 2.42458850e-01 -8.87332931e-02 8.08220267e-01
-3.01334351e-01 -1.10726804e-01 1.74038500e-01 -2.67492812e-02
7.75663078e-01 -3.05739671e-01 2.99904674e-01 -9.70335007e-01
1.44324571e-01 -3.92480195e-01 -4.47863728e-01 -3.12703475e-02
-4.02020752e-01 -9.07378316e-01 -1.63931549e-01 -1.78382051e+00
1.56429738e-01 -3.43911678e-01 -5.02757907e-01 4.27801877e-01
-2.98953563e-01 5.71011484e-01 1.85758531e-01 -1.22132942e-01
2.63347745e-01 -1.92475498e-01 1.84616506e+00 -2.36846298e-01
-7.27146208e-01 -2.49066502e-02 -5.11833012e-01 6.47950649e-01
1.00373399e+00 -2.78375596e-01 -3.78941894e-01 2.94017673e-01
-5.19852340e-01 1.10149108e-01 3.52585524e-01 -8.55918944e-01
1.90562621e-01 -5.13200462e-01 8.13767314e-01 -5.72797835e-01
1.73246875e-01 -9.71094549e-01 1.09937511e-01 8.58498871e-01
-9.89795625e-02 -1.87348440e-01 2.28256568e-01 5.95690012e-02
-5.72456121e-02 -5.40086865e-01 1.36117995e+00 -2.74417698e-01
-5.69854617e-01 5.95893385e-03 -4.77160871e-01 -4.88935739e-01
1.34016633e+00 -1.10677755e+00 -2.33949095e-01 -1.06623515e-01
-9.23077762e-01 -3.64694834e-01 4.36949432e-01 -1.89089283e-01
8.49390626e-01 -1.04909992e+00 -5.27136505e-01 2.74524391e-01
-3.06109339e-01 -2.97672093e-01 6.09537125e-01 1.41227341e+00
-1.22899902e+00 1.13701813e-01 -7.94666827e-01 -5.78898311e-01
-1.80809963e+00 1.24383725e-01 8.62034738e-01 -2.74385422e-01
-2.68585742e-01 6.62558913e-01 2.95151845e-02 -6.36247471e-02
-8.06434676e-02 -4.01639700e-01 -2.53952622e-01 -2.60102630e-01
3.90861183e-01 7.43614137e-01 9.87173989e-02 -5.33116460e-01
-1.21153608e-01 4.75987852e-01 -1.32685348e-01 -2.18162850e-01
1.20789123e+00 -7.53564388e-02 -3.94914061e-01 2.26675048e-01
1.25425553e+00 -1.31890908e-01 -1.12909257e+00 2.55866289e-01
-3.95305425e-01 -6.28353953e-01 2.31867239e-01 -1.32596958e+00
-1.15904677e+00 8.15716565e-01 1.37685311e+00 2.64217407e-01
1.59448647e+00 -5.13229966e-01 6.17648661e-01 -8.14886391e-02
-3.49943906e-01 -1.68354881e+00 4.86304276e-02 -2.17030384e-02
6.61965251e-01 -9.03126061e-01 1.13789933e-02 -4.42208081e-01
-8.56697261e-01 1.31688559e+00 7.37557530e-01 -3.88182759e-01
2.88954824e-01 3.56696129e-01 4.83498424e-01 -1.30918279e-01
-1.75913304e-01 -1.00434944e-01 3.98199022e-01 8.82637262e-01
2.59440482e-01 -8.36341009e-02 -9.44895327e-01 1.22954778e-01
1.15828924e-01 7.27313012e-02 5.66385150e-01 8.96430373e-01
-5.57339489e-01 -7.00560451e-01 -8.05390596e-01 6.61093175e-01
-6.05475366e-01 1.10198893e-01 -3.57413799e-01 1.06959045e+00
5.36312699e-01 9.83788788e-01 2.46124730e-01 -1.24453112e-01
2.13665515e-01 -1.90640166e-01 8.15246642e-01 -9.88852531e-02
-7.46088862e-01 4.31146562e-01 -2.26251081e-01 -2.19775736e-01
-4.42134053e-01 -2.25976259e-01 -1.41992223e+00 -3.36532921e-01
-4.57343429e-01 -3.23335007e-02 1.01470447e+00 8.23712528e-01
-4.44202960e-01 5.92061460e-01 6.51584029e-01 -3.21292877e-01
-2.46952578e-01 -8.80574048e-01 -9.34026420e-01 2.35604823e-01
1.80472568e-01 -3.85863841e-01 -3.94069433e-01 3.15396905e-01] | [15.543158531188965, -2.991095542907715] |
8a3cd836-b2c6-440f-9782-567e1a49ec48 | a-deep-active-contour-model-for-delineating | 2307.03461 | null | https://arxiv.org/abs/2307.03461v1 | https://arxiv.org/pdf/2307.03461v1.pdf | A Deep Active Contour Model for Delineating Glacier Calving Fronts | Choosing how to encode a real-world problem as a machine learning task is an important design decision in machine learning. The task of glacier calving front modeling has often been approached as a semantic segmentation task. Recent studies have shown that combining segmentation with edge detection can improve the accuracy of calving front detectors. Building on this observation, we completely rephrase the task as a contour tracing problem and propose a model for explicit contour detection that does not incorporate any dense predictions as intermediate steps. The proposed approach, called ``Charting Outlines by Recurrent Adaptation'' (COBRA), combines Convolutional Neural Networks (CNNs) for feature extraction and active contour models for the delineation. By training and evaluating on several large-scale datasets of Greenland's outlet glaciers, we show that this approach indeed outperforms the aforementioned methods based on segmentation and edge-detection. Finally, we demonstrate that explicit contour detection has benefits over pixel-wise methods when quantifying the models' prediction uncertainties. The project page containing the code and animated model predictions can be found at \url{https://khdlr.github.io/COBRA/}. | ['Xiao Xiang Zhu', 'Sébastien Lefèvre', 'Mirko Scheinert', 'Erik Loebel', 'Lichao Mou', 'Konrad Heidler'] | 2023-07-07 | null | null | null | null | ['edge-detection', 'contour-detection'] | ['computer-vision', 'computer-vision'] | [ 1.15684174e-01 4.99733180e-01 9.54753906e-02 -4.79796469e-01
-9.45716739e-01 -7.72852361e-01 6.89537823e-01 3.20870906e-01
-4.14047688e-01 4.15373385e-01 8.09744745e-02 -8.49484861e-01
6.19656630e-02 -9.95790243e-01 -6.66260362e-01 -6.91716254e-01
-2.68441170e-01 4.85226780e-01 2.49871075e-01 -2.28937924e-01
2.34877095e-01 6.73689187e-01 -1.28662109e+00 -1.17954589e-01
1.17699635e+00 9.99038756e-01 -1.35418763e-02 7.03774452e-01
-2.30343863e-01 6.65434122e-01 -3.40735167e-01 -1.01323627e-01
3.75279307e-01 -2.44373724e-01 -5.47866225e-01 -8.39440972e-02
5.04273176e-01 2.18033288e-02 2.14312077e-01 1.00583565e+00
3.38787854e-01 1.00939229e-01 6.07716680e-01 -8.64802837e-01
-9.39688608e-02 6.21128976e-01 -4.90629256e-01 2.23203465e-01
-4.12533224e-01 7.68113509e-02 1.13998401e+00 -8.62515509e-01
4.16045874e-01 8.87110412e-01 1.32375038e+00 2.88190916e-02
-1.34904087e+00 -5.21870255e-01 1.08555332e-01 -2.77735498e-02
-1.40510774e+00 -4.44118321e-01 8.43681574e-01 -8.53114486e-01
8.02611351e-01 3.77231538e-01 9.01585579e-01 2.41726562e-01
2.69600935e-02 9.25617635e-01 9.05974209e-01 -4.46640104e-01
5.72578311e-01 -2.22661033e-01 3.50018263e-01 6.21821582e-01
6.07754350e-01 4.49893683e-01 -2.23861560e-01 -9.42729339e-02
6.13257170e-01 -3.25987160e-01 -4.63442504e-01 -1.05881229e-01
-5.81819057e-01 1.08514404e+00 6.53002501e-01 1.02724701e-01
-3.49501222e-01 4.68076289e-01 1.24257118e-01 4.28667245e-03
1.07203197e+00 5.17058551e-01 -4.44522530e-01 2.20565498e-01
-1.67636061e+00 5.00453353e-01 8.49495232e-01 6.59137189e-01
1.04860115e+00 2.50208437e-01 1.42996401e-01 6.38531744e-01
8.31660569e-01 6.81487381e-01 -1.10447407e-01 -9.20524299e-01
1.63538143e-01 8.27390909e-01 3.03919286e-01 -8.41978431e-01
-6.75072968e-01 -5.66168129e-01 -7.26839006e-01 9.02833760e-01
4.92064655e-01 -6.37924016e-01 -1.16126513e+00 1.16120267e+00
1.34627447e-01 8.26556161e-02 7.67690092e-02 8.78862083e-01
8.62985075e-01 7.27293015e-01 3.99708331e-01 1.05309360e-01
1.09758520e+00 -7.82491565e-01 -6.84155524e-01 -6.38630807e-01
6.34391427e-01 -5.03592432e-01 5.53196490e-01 1.41193002e-01
-7.97805727e-01 -4.88869011e-01 -1.37549162e+00 8.13515708e-02
-5.57815790e-01 2.63217092e-01 7.90008187e-01 5.60403764e-01
-1.24610913e+00 9.23331022e-01 -1.14909959e+00 -8.90539885e-02
5.60233474e-01 -3.68984640e-02 9.42812040e-02 5.54267704e-01
-1.31015277e+00 9.49187994e-01 3.43168855e-01 7.80478001e-01
-7.45978177e-01 -6.18019581e-01 -1.11326897e+00 1.15401484e-01
2.31836215e-01 -3.02870244e-01 1.27147758e+00 -1.26752579e+00
-1.22788727e+00 7.99611390e-01 4.25819829e-02 -8.65554631e-01
8.30436945e-01 -4.22890067e-01 -4.82251868e-02 9.07142013e-02
-1.32312715e-01 6.37550771e-01 7.76199162e-01 -1.48138499e+00
-7.38366663e-01 -1.95774958e-01 -2.37757936e-01 2.52309710e-01
2.97425717e-01 -1.47217289e-01 -1.46325156e-01 -8.06750417e-01
4.69024986e-01 -8.09133887e-01 -5.84882438e-01 3.18893760e-01
-1.13790944e-01 5.30959144e-02 6.41100347e-01 -9.41685200e-01
1.21188700e+00 -1.93240273e+00 -5.31988256e-02 3.04565370e-01
1.53747752e-01 4.48240727e-01 4.79263037e-01 3.52853507e-01
6.98717982e-02 4.36070979e-01 -1.28710723e+00 -4.80792224e-01
9.94850919e-02 1.86045974e-01 -3.98895085e-01 6.12694919e-01
3.66178632e-01 9.41855431e-01 -6.39524221e-01 -4.14227515e-01
3.35128874e-01 4.60222512e-01 -6.11673333e-02 1.18011393e-01
-4.60921317e-01 3.99757504e-01 -2.33457312e-01 4.64617997e-01
1.11778021e+00 -9.96885598e-02 1.89764425e-01 1.98408112e-01
-6.22994840e-01 -6.47996143e-02 -1.17780662e+00 1.45916271e+00
-2.49185085e-01 9.65675652e-01 2.29429364e-01 -8.09878528e-01
1.36263335e+00 2.50663280e-01 2.38600239e-01 -5.06772995e-01
-1.70883998e-01 4.86670703e-01 -1.37347847e-01 -3.46389264e-01
5.77446699e-01 -2.34988272e-01 3.14371735e-02 1.26291707e-01
-2.06340730e-01 -4.77402896e-01 -1.44900560e-01 -1.48533627e-01
5.88709772e-01 8.36033165e-01 2.76106894e-01 -8.49305391e-01
4.01732653e-01 7.35726655e-01 8.00551951e-01 7.20683634e-01
-3.47589344e-01 7.01230526e-01 3.99301678e-01 -7.19995201e-01
-1.04759061e+00 -7.16003299e-01 -5.17593086e-01 8.42259228e-01
1.67088598e-01 -1.50054216e-01 -7.99127400e-01 -3.30416709e-01
1.85835794e-01 8.15260530e-01 -8.38092983e-01 3.10662091e-01
-6.05550230e-01 -1.20772660e+00 8.11293364e-01 7.51519620e-01
7.17538774e-01 -1.05555093e+00 -1.13564515e+00 2.75560230e-01
9.77846235e-02 -4.85054880e-01 1.05621658e-01 5.32555699e-01
-1.10032511e+00 -9.19588327e-01 -7.26885319e-01 -5.61548829e-01
4.81556714e-01 -2.41552085e-01 1.20532668e+00 2.57996559e-01
-1.61742657e-01 -7.44454786e-02 -5.23547292e-01 -4.70181584e-01
-3.27963948e-01 9.88868698e-02 -8.01555991e-01 -2.73790807e-01
1.99044198e-01 -3.96318585e-01 -7.91247845e-01 5.20266443e-02
-8.71355772e-01 4.76629108e-01 4.66709495e-01 5.55122793e-01
5.28190374e-01 -2.67033547e-01 4.46094543e-01 -1.07529283e+00
2.15515882e-01 -3.66111219e-01 -9.91099834e-01 2.60925442e-01
-8.52159202e-01 1.10650122e-01 2.11186945e-01 4.48739260e-01
-1.30620039e+00 5.81069827e-01 -5.98511875e-01 -9.21481661e-03
-4.81062680e-01 8.29345942e-01 1.06442727e-01 -5.27674798e-03
8.37716103e-01 6.94060698e-02 -1.20465659e-01 -6.71793640e-01
4.11990494e-01 6.21063590e-01 4.35502261e-01 -2.52652854e-01
8.16324294e-01 6.86515033e-01 -1.23695180e-01 -8.40069175e-01
-6.75500154e-01 -6.49233639e-01 -8.96522641e-01 -5.80344498e-01
8.44995558e-01 -1.05456519e+00 -2.22345546e-01 7.29471326e-01
-1.04684675e+00 -1.07965922e+00 -1.77668482e-01 1.99247926e-01
-4.78401005e-01 2.59894967e-01 -3.07289094e-01 -1.18918502e+00
-7.02654898e-01 -7.55786359e-01 1.00055981e+00 5.76984763e-01
3.21325799e-03 -1.36163640e+00 3.41216147e-01 2.15479154e-02
6.71517611e-01 7.49481618e-01 4.90713477e-01 -5.05694985e-01
-5.22898793e-01 -8.73797238e-02 -1.85624316e-01 1.87419519e-01
-1.05621904e-01 3.09611619e-01 -1.32438111e+00 -6.39489815e-02
-6.00497611e-02 1.42321452e-01 1.46002138e+00 7.00073779e-01
5.06710470e-01 -2.32378468e-01 -2.72591084e-01 9.55902517e-01
1.78966010e+00 3.02501351e-01 7.88933814e-01 5.73530614e-01
4.78515536e-01 7.39775777e-01 6.16761744e-01 4.70305890e-01
1.50870368e-01 2.08781585e-01 4.87471074e-01 -6.83923900e-01
-4.90492322e-02 -1.66108981e-02 -3.57949734e-02 2.42589101e-01
-3.71579170e-01 -1.76834270e-01 -1.67746365e+00 5.99715769e-01
-2.10435128e+00 -9.04915810e-01 -6.36494398e-01 2.02624559e+00
6.37398839e-01 2.20949218e-01 -3.38301778e-01 2.05544159e-01
6.31956577e-01 4.49966043e-01 -4.07776535e-01 -2.71201164e-01
-2.93401301e-01 1.81414872e-01 1.08056235e+00 1.11639774e+00
-1.68015766e+00 1.23137307e+00 5.31601048e+00 4.48798388e-01
-1.23012114e+00 6.75588772e-02 8.63804758e-01 8.18204463e-01
-3.11474144e-01 4.01430786e-01 -8.04114163e-01 2.06622500e-02
8.29622269e-01 -9.79744494e-02 1.85501426e-02 7.62288094e-01
5.98477006e-01 -4.50344205e-01 -4.64165181e-01 3.03982407e-01
-2.10270375e-01 -1.42603230e+00 -3.86901677e-01 -2.94073373e-01
6.69422150e-01 1.89719170e-01 -3.35319161e-01 -2.14950815e-02
5.49760818e-01 -9.79973853e-01 1.13878572e+00 9.55231369e-01
4.82990056e-01 -4.36390907e-01 6.94055140e-01 1.66182712e-01
-1.52442276e+00 1.23834588e-01 -1.76074162e-01 -1.41034558e-01
2.38916218e-01 5.10757625e-01 -7.58608282e-01 5.96175492e-01
7.28751779e-01 8.70281398e-01 -7.96313286e-01 1.57517540e+00
-4.82948452e-01 1.14452899e+00 -4.81677949e-01 4.48495835e-01
3.43632072e-01 -3.18336040e-01 5.43801785e-01 1.42365086e+00
6.78063259e-02 2.83615813e-02 -1.23891987e-01 1.22284758e+00
3.47868860e-01 1.82200626e-01 -3.68929774e-01 2.66309828e-01
2.22300440e-01 1.24328136e+00 -1.16566551e+00 -3.28533024e-01
-1.45792991e-01 5.85350633e-01 -2.05529816e-02 3.97168398e-01
-8.20531726e-01 -5.07255256e-01 3.51737380e-01 9.63944793e-02
2.96888471e-01 -3.61789644e-01 -8.66183877e-01 -6.51745856e-01
-2.76038915e-01 -1.80140927e-01 4.05282706e-01 -8.36946189e-01
-7.41385460e-01 6.67217076e-01 8.54315534e-02 -1.37284160e+00
-3.11546195e-02 -6.18384719e-01 -9.23260093e-01 1.02180314e+00
-1.90820181e+00 -1.33006620e+00 -7.29957938e-01 -2.04172671e-01
5.37608743e-01 4.18380797e-01 7.40010738e-01 9.98635888e-02
-3.96567881e-01 -1.92076191e-02 4.53118742e-01 4.10693109e-01
2.83040136e-01 -1.61197519e+00 8.36929202e-01 1.22837365e+00
-1.52081802e-01 1.27061114e-01 7.77762592e-01 -9.70860839e-01
-8.25931430e-01 -1.45329177e+00 7.51324773e-01 -2.46161576e-02
5.81543863e-01 -2.97323316e-01 -1.25156140e+00 8.59928787e-01
1.88033000e-01 9.81545225e-02 4.50338006e-01 -2.60480076e-01
1.61347941e-01 7.47472048e-02 -8.52092803e-01 2.71283507e-01
7.64974833e-01 -1.14543311e-01 -6.28509223e-01 1.16753563e-01
3.31048876e-01 -4.26232994e-01 -8.06233883e-01 7.19944000e-01
3.17003459e-01 -1.01725018e+00 6.08452201e-01 -8.99114832e-02
4.12911415e-01 -4.74648714e-01 6.01135232e-02 -1.15829527e+00
-2.01404899e-01 -4.23298955e-01 1.78448260e-01 1.01700139e+00
7.73258984e-01 -6.96410477e-01 8.02294731e-01 5.07024288e-01
-7.03035593e-01 -4.37312067e-01 -9.71766591e-01 -4.04495507e-01
5.46128154e-01 -4.97114837e-01 3.87167126e-01 7.67050087e-01
-2.54400283e-01 -2.09006295e-01 -3.79739539e-03 5.94294190e-01
6.40170991e-01 4.56386685e-01 3.31709892e-01 -1.71978295e+00
9.07499865e-02 -6.37182415e-01 -1.78721085e-01 -8.40353310e-01
-4.67018224e-02 -7.56739914e-01 5.58800578e-01 -1.81953692e+00
-4.24825430e-01 -8.34695160e-01 1.06396519e-01 8.82241428e-01
-2.34824255e-01 2.10903034e-01 3.13398093e-02 4.88740981e-01
-4.52057719e-02 5.70079088e-01 1.00522399e+00 -1.00303218e-01
-2.59857059e-01 7.59148076e-02 -1.39096186e-01 8.71663153e-01
1.08467209e+00 -4.67736453e-01 8.63383934e-02 -3.46075416e-01
3.64440560e-01 1.54374570e-01 4.18335289e-01 -1.23320937e+00
3.64548922e-01 -7.76029900e-02 1.90835446e-01 -7.39120841e-01
8.16122144e-02 -7.74515986e-01 4.53783423e-01 5.79329908e-01
-2.28937387e-01 -3.22046310e-01 6.74594104e-01 4.83842343e-01
-2.39786506e-01 -5.12486100e-01 9.35686409e-01 -2.81046599e-01
-1.08750296e+00 1.12543851e-02 -6.21142387e-01 -1.13078557e-01
7.40149140e-01 -2.46322498e-01 -4.32652503e-01 -7.90819377e-02
-9.45164561e-01 3.15722883e-01 5.45339346e-01 7.88928419e-02
4.37859684e-01 -6.29639924e-01 -9.19202328e-01 1.65460676e-01
2.25657895e-02 5.44187367e-01 2.46564090e-01 8.33341599e-01
-1.40775549e+00 2.44356483e-01 3.55451591e-02 -6.62744701e-01
-8.56926918e-01 -2.98044272e-02 1.10405076e+00 -1.04989409e-01
-9.53397989e-01 7.72135198e-01 -2.64665298e-02 -3.23944271e-01
-1.18027359e-01 -5.62723100e-01 -4.51815844e-01 4.38187420e-01
1.18828267e-01 1.39546260e-01 2.35748738e-01 -8.13874900e-01
-1.41110688e-01 5.85587382e-01 5.73665917e-01 -2.02807888e-01
1.48663092e+00 -1.30097508e-01 1.06403530e-01 5.23320317e-01
6.12722099e-01 -3.54437053e-01 -1.66502428e+00 2.42671333e-02
5.73829234e-01 -2.20954135e-01 5.41324258e-01 -9.65619743e-01
-1.28690767e+00 9.43063378e-01 8.27796996e-01 1.16005681e-01
9.54219043e-01 -2.26615518e-01 1.61079779e-01 2.91470647e-01
-7.01114833e-02 -1.03153872e+00 -6.24546111e-01 4.32418555e-01
1.05623865e+00 -1.39829409e+00 2.97048800e-02 -3.58312875e-01
-7.12526977e-01 1.28526604e+00 3.29151362e-01 -2.62884140e-01
9.84637201e-01 3.75842273e-01 6.23197556e-01 -3.70415300e-01
-3.43938112e-01 -4.74910468e-01 1.25059515e-01 1.87051430e-01
3.63860011e-01 3.31002504e-01 -2.82123268e-01 4.68385011e-01
-7.24325627e-02 8.41321610e-03 4.85699058e-01 1.02986896e+00
-8.34996283e-01 -7.57144332e-01 -5.90613067e-01 3.77893925e-01
-1.78888217e-01 -3.61534417e-01 -2.73042262e-01 6.90094769e-01
2.00253263e-01 6.29520535e-01 4.44869459e-01 -7.99215734e-02
-8.78500659e-03 2.33338818e-01 -2.50747859e-01 -5.16151786e-01
-7.69335628e-01 6.62462935e-02 1.46644801e-01 5.20539796e-03
-5.89328766e-01 -7.11756349e-01 -1.49959195e+00 3.75248715e-02
-3.10721815e-01 3.46429467e-01 6.57003164e-01 7.22097754e-01
5.20062223e-02 4.37230408e-01 1.82028919e-01 -1.06174409e+00
-1.20671540e-01 -1.16131330e+00 -7.05995202e-01 1.46941602e-01
3.96807551e-01 -6.79396629e-01 -6.91428721e-01 3.85115355e-01] | [9.554694175720215, -1.4818305969238281] |
7aabeb03-7673-4180-966b-b78129894abc | mci-net-multi-scale-context-integrated | null | null | https://www.sciencedirect.com/science/article/pii/S0045790622003408 | https://www.sciencedirect.com/science/article/pii/S0045790622003408 | Mci-net: multi-scale context integrated network for liver ct image segmentation | Owing to the various object scales and high similarity with the surrounding organs (e.g., kidney, stomach, and spleen), it is difficult to accurately segment the liver region from the abdominal computed tomography images. In this study, we propose a multi-scale context integration network called MCI-Net for liver image segmentation. Specifically, we first design a simplified residual module to prevent network degradation. Given the scale variability of objects, we propose a multi-scale context extraction module by combining four cascaded branches of hybrid dilated convolutions to capture broader and deeper features. In addition, we introduce an external attention mechanism based on two external, learnable and shared memory units, which helps to perceive the most discriminative information and suppress redundant features. Finally, we provide a boundary correction block to further improve the localization ability of boundary information. Extensive experiments on two liver CT image benchmark datasets qualitatively and quantitatively illustrate that our method is effective in improving liver segmentation accuracy and outperforms several state-of-the-art methods. | ['Jubai An', 'Weidong Zhang', 'Feng Shao', 'Xipeng Pan', 'Xiwang Xie'] | 2023-05-03 | null | null | null | computers-and-electrical-engineering-2023-5 | ['2d-semantic-segmentation', 'liver-segmentation'] | ['computer-vision', 'medical'] | [-1.79897919e-01 -4.63910289e-02 -1.58635795e-01 -4.31401879e-01
-4.26436633e-01 -3.55300605e-01 1.14042036e-01 1.89149871e-01
-3.25343221e-01 3.58000040e-01 3.54225993e-01 -1.93432719e-01
1.15438499e-01 -5.22567272e-01 -4.25259382e-01 -8.12698543e-01
-1.64517134e-01 -1.09193549e-01 4.80496436e-01 1.39206171e-01
-1.49445474e-01 5.28436840e-01 -5.27299762e-01 2.38562077e-01
1.05009019e+00 1.03896546e+00 4.67122078e-01 3.03375542e-01
-1.60041854e-01 7.45353878e-01 -3.10854763e-01 1.84504181e-01
1.77534580e-01 -6.16070032e-01 -8.06980848e-01 3.16287100e-01
1.92918703e-02 -4.93714035e-01 -4.39239770e-01 1.26798165e+00
5.44150293e-01 8.26743692e-02 3.89663100e-01 -5.50171435e-01
-7.27899671e-01 6.81543171e-01 -7.97176003e-01 6.69493854e-01
-4.78782132e-02 2.72494107e-01 5.53719580e-01 -9.08411324e-01
3.34222198e-01 9.15081501e-01 6.47510409e-01 2.54321873e-01
-1.02276897e+00 -6.49130106e-01 3.97317410e-01 -2.58263797e-01
-1.41601324e+00 8.59820843e-02 8.00218701e-01 -1.38648301e-01
3.98195386e-01 2.26137653e-01 8.35483968e-01 6.03422761e-01
4.01073784e-01 9.54033256e-01 8.81935477e-01 -1.15380391e-01
-8.66990983e-02 -7.98792690e-02 -9.98911262e-03 1.06393147e+00
3.24171543e-01 -5.02664559e-02 2.95289516e-01 -8.98404866e-02
1.38603592e+00 4.15705651e-01 -5.91388583e-01 -6.27516627e-01
-1.58416629e+00 8.52592707e-01 1.28961182e+00 5.89412451e-01
-4.10159171e-01 1.48785889e-01 3.82532656e-01 -2.11269945e-01
3.43066156e-01 2.33839974e-01 -5.33195674e-01 7.03414083e-01
-7.80249774e-01 -4.33857709e-01 6.97876275e-01 8.05828273e-01
5.40389419e-01 -9.09977183e-02 -5.60946822e-01 6.89999938e-01
4.85478699e-01 7.35264942e-02 6.94488823e-01 -4.25613165e-01
1.51228860e-01 8.60651553e-01 -3.45260352e-01 -7.35975862e-01
-8.26037169e-01 -8.85488987e-01 -1.21171141e+00 -1.22732691e-01
1.99953780e-01 -2.04690486e-01 -1.19747603e+00 1.50792277e+00
5.67489088e-01 5.99011540e-01 -2.02687785e-01 1.44458961e+00
1.19616008e+00 3.09309483e-01 4.81106669e-01 -1.04898736e-01
1.64786160e+00 -1.41286790e+00 -5.00284016e-01 -2.14013264e-01
5.15254080e-01 -7.09314048e-01 8.60329032e-01 -2.78106689e-01
-9.59652126e-01 -4.11764771e-01 -9.21316028e-01 -7.76302963e-02
-5.95153756e-02 4.11880255e-01 8.52312803e-01 1.94562674e-01
-6.72147989e-01 4.55861658e-01 -1.21520472e+00 -2.21475959e-01
6.05115414e-01 3.19894016e-01 -5.43991029e-02 1.51902782e-02
-1.03995550e+00 6.86518669e-01 4.63910162e-01 2.40895435e-01
-8.66625786e-01 -6.59651935e-01 -1.11207819e+00 2.42328838e-01
3.22224140e-01 -9.66417968e-01 1.09405684e+00 -1.03841984e+00
-1.48681068e+00 6.33798778e-01 1.64758250e-01 -1.34197176e-01
4.55615968e-01 1.81427985e-01 -8.62788409e-02 6.03757083e-01
2.67002080e-02 8.19290638e-01 5.58693111e-01 -1.04650545e+00
-3.03417444e-01 -1.86042979e-01 -3.87337096e-02 3.16551447e-01
-7.44696781e-02 2.35501900e-02 -7.37112343e-01 -1.11141932e+00
4.41638857e-01 -8.09917927e-01 -7.54546285e-01 2.42570147e-01
-3.47722203e-01 3.60517651e-02 6.64445639e-01 -9.47924614e-01
1.22629392e+00 -2.22736287e+00 1.80868313e-01 2.17319220e-01
4.32643592e-01 5.04529513e-02 -5.16362749e-02 -4.13141996e-01
-2.23875806e-01 1.46196485e-01 -3.35147262e-01 -2.00572252e-01
-4.02506739e-01 4.37611103e-01 2.85272658e-01 5.98098397e-01
1.90696761e-01 1.30541527e+00 -9.95494604e-01 -8.22632074e-01
3.10673714e-01 5.11104167e-01 -4.39700603e-01 4.22501564e-02
3.11860088e-02 9.64133084e-01 -7.79037774e-01 8.47863913e-01
7.00134873e-01 -8.26061606e-01 9.17112902e-02 -4.51126635e-01
-5.93010001e-02 1.11230217e-01 -8.59446466e-01 2.14823055e+00
-4.90396708e-01 1.27352357e-01 3.57570112e-01 -8.34490955e-01
5.28445125e-01 3.55988771e-01 5.54339290e-01 -5.39249897e-01
4.30926830e-01 1.51305348e-01 1.07602768e-01 -4.85157430e-01
-9.43752006e-03 -3.81430574e-02 6.72188774e-02 2.72510201e-01
4.23589051e-02 5.11446670e-02 -5.57508767e-02 7.77364746e-02
8.03579330e-01 -1.15943864e-01 5.25669277e-01 -6.38674140e-01
7.44700789e-01 -2.64383346e-01 9.50072467e-01 3.72851849e-01
-6.06263995e-01 8.52977157e-01 3.76473606e-01 -6.80292189e-01
-6.59877419e-01 -8.79296780e-01 -2.17139751e-01 8.13733995e-01
7.00685799e-01 -1.48781454e-02 -8.08045864e-01 -1.14792323e+00
-4.37019095e-02 -4.08946313e-02 -7.68239558e-01 -1.10928580e-01
-8.94824028e-01 -9.50684071e-01 2.36424252e-01 8.83806050e-01
7.24294901e-01 -9.41848099e-01 -5.33681393e-01 2.72094309e-01
-2.39683762e-01 -8.54766607e-01 -1.06265545e+00 2.58112967e-01
-1.11918485e+00 -1.14787710e+00 -1.02199173e+00 -1.27414894e+00
1.08561313e+00 5.43593407e-01 1.14529133e+00 5.66965103e-01
-5.61149895e-01 -8.26004446e-02 -1.28719568e-01 6.34989282e-03
-2.34486610e-02 1.61865383e-01 -4.94332641e-01 -2.58544207e-01
3.58773698e-03 -3.58996034e-01 -1.26031744e+00 3.57544184e-01
-8.94638062e-01 2.13194817e-01 1.03962326e+00 1.15363359e+00
7.74379134e-01 -1.61757424e-01 4.60228980e-01 -6.87209606e-01
3.64481628e-01 -5.45908272e-01 -4.86662775e-01 4.14693177e-01
-3.50942999e-01 -7.11554289e-02 5.49683094e-01 -5.96314013e-01
-9.75855291e-01 3.17675442e-01 5.82890846e-02 -4.27864373e-01
-6.73566684e-02 5.09978414e-01 1.07522398e-01 -5.46216667e-01
2.92215437e-01 1.60218418e-01 -7.01436773e-03 -3.69208843e-01
1.73060879e-01 1.53874919e-01 4.04168755e-01 -3.06033760e-01
4.03445482e-01 5.00535309e-01 -1.62297755e-01 -3.10265243e-01
-8.17795336e-01 -4.58809763e-01 -5.06066024e-01 1.63775921e-01
1.08842576e+00 -1.08720291e+00 -4.53095168e-01 3.38034451e-01
-9.25815761e-01 -4.05772001e-01 -2.79757440e-01 7.13046372e-01
-2.20878810e-01 5.60936928e-01 -1.33111846e+00 1.58267785e-02
-6.68694675e-01 -1.54529774e+00 1.00243974e+00 7.78079927e-01
2.76673138e-01 -1.08753705e+00 -2.40614980e-01 -8.68698768e-03
6.61783814e-01 2.54857987e-01 8.04191649e-01 -6.07155979e-01
-9.08198297e-01 1.98227540e-01 -7.53046095e-01 2.98497468e-01
3.33590806e-01 -3.52197260e-01 -3.79161865e-01 -5.02485514e-01
7.28232265e-02 8.22568461e-02 1.08076656e+00 7.22100198e-01
1.37934887e+00 -2.08711132e-01 -5.33339322e-01 1.14684594e+00
1.32729983e+00 5.69609553e-02 1.88298479e-01 -2.36385223e-02
6.51183307e-01 5.65779433e-02 2.71117449e-01 2.16676623e-01
3.86161417e-01 9.03549492e-02 4.48695570e-01 -8.47408354e-01
-4.43645209e-01 8.18062276e-02 -2.96094805e-01 9.99012589e-01
2.40699783e-01 2.55169809e-01 -7.32162118e-01 5.67653239e-01
-1.69454384e+00 -3.70949090e-01 2.31823191e-01 1.81543720e+00
7.91537583e-01 -5.67419790e-02 -1.61639065e-01 -7.17642963e-01
7.81686902e-01 1.81567743e-02 -5.80049574e-01 2.99108177e-01
1.06681988e-01 1.25970632e-01 5.69036365e-01 3.52376491e-01
-1.57822168e+00 6.16722941e-01 6.11171055e+00 6.83678269e-01
-1.32034695e+00 2.73210943e-01 9.79049504e-01 1.56006053e-01
-1.45149037e-01 -1.03008188e-02 -2.27536529e-01 5.66226900e-01
-4.06408533e-02 -2.87051243e-03 2.42091343e-01 8.41348052e-01
-5.27336225e-02 -7.72750601e-02 -9.08302963e-01 7.45582819e-01
6.82475716e-02 -1.25956488e+00 -1.29901454e-01 -1.32385179e-01
9.09900486e-01 3.57504815e-01 -1.88089162e-01 2.82398403e-01
1.33588776e-01 -7.87507594e-01 2.59016424e-01 4.75306213e-01
5.96610785e-01 -5.99189997e-01 9.24077094e-01 2.18651503e-01
-1.47228086e+00 -1.22950472e-01 -3.66537869e-01 3.63347560e-01
-1.48231998e-01 5.57161450e-01 -5.18535852e-01 6.70965672e-01
5.33472657e-01 6.62361979e-01 -6.91499293e-01 1.42807114e+00
-3.46863687e-01 3.73473823e-01 -3.71794403e-01 1.64780453e-01
3.98916841e-01 -1.89304054e-01 3.48276883e-01 1.46567225e+00
8.92089084e-02 4.12889391e-01 6.89068019e-01 9.36871648e-01
-2.14215025e-01 1.79945856e-01 -6.73879832e-02 4.39633518e-01
1.51205584e-01 1.75504768e+00 -1.24057424e+00 -5.39406300e-01
-6.43140912e-01 1.04671609e+00 2.09416643e-01 4.78193969e-01
-9.33365047e-01 -2.93879718e-01 2.27971509e-01 -3.74929756e-01
2.63308823e-01 -3.89868096e-02 -3.43414307e-01 -1.56565630e+00
-1.69410214e-01 -5.83339214e-01 5.18618226e-01 -3.60177904e-01
-1.24959922e+00 6.71863973e-01 -3.68184596e-01 -1.17787242e+00
3.62441301e-01 -2.69808561e-01 -9.40948248e-01 8.89060497e-01
-1.73778296e+00 -1.34235227e+00 -7.59299755e-01 4.60662842e-01
5.18536389e-01 2.77359962e-01 5.68933249e-01 4.09382880e-01
-6.20767415e-01 3.74419332e-01 -1.61158234e-01 6.80273235e-01
6.84081078e-01 -1.25821674e+00 1.19622663e-01 8.15706849e-01
-3.97904396e-01 8.16743076e-01 1.70654207e-02 -6.64338529e-01
-1.01055419e+00 -1.27794957e+00 2.55839080e-01 1.58469379e-01
5.08684754e-01 -6.73666596e-02 -1.03275549e+00 7.45370746e-01
2.58285344e-01 9.03013647e-01 5.87404251e-01 -3.89320374e-01
-1.58373173e-02 -4.66984622e-02 -1.21044803e+00 4.51547325e-01
7.50003457e-01 -2.21288815e-01 -7.47550428e-01 3.47605199e-01
8.77218127e-01 -8.48065138e-01 -1.00434637e+00 7.16323078e-01
3.63655061e-01 -6.95053279e-01 1.15756118e+00 -2.93674737e-01
2.56315440e-01 -4.01107222e-01 2.29797974e-01 -1.24635983e+00
-6.89487398e-01 -2.29600772e-01 -5.38452603e-02 7.60329723e-01
9.26156342e-02 -6.17193162e-01 7.38889694e-01 4.38856900e-01
-4.16033894e-01 -1.15228903e+00 -7.55313218e-01 -4.20553863e-01
1.81071267e-01 4.25882846e-01 5.85819840e-01 1.13644087e+00
-7.45631456e-02 1.64908528e-01 1.43493488e-01 2.96321005e-01
6.23700261e-01 6.89711094e-01 1.78462058e-01 -8.34367633e-01
-2.61743993e-01 -6.52815282e-01 -2.40724578e-01 -1.43670011e+00
-1.66598856e-01 -1.01005447e+00 -8.46337602e-02 -1.55532742e+00
6.66377783e-01 -6.99766278e-01 -8.29305232e-01 5.08473933e-01
-6.19583189e-01 2.80207843e-01 1.36713788e-01 2.53767192e-01
-8.14121008e-01 4.09101725e-01 1.91606808e+00 -1.77877694e-01
-1.56368703e-01 -1.22216962e-01 -6.37138307e-01 9.09085393e-01
7.69014359e-01 -2.86578685e-01 -7.42015243e-02 -4.88805711e-01
-4.10745442e-01 3.11287612e-01 4.56614792e-01 -8.30286741e-01
4.12633419e-01 -9.67855006e-02 9.38570201e-01 -5.78953147e-01
-2.35951826e-01 -8.65387857e-01 -1.81172833e-01 8.41757894e-01
-2.88614482e-01 -1.89390510e-01 1.79366305e-01 4.29516107e-01
-3.72135490e-01 -6.46394044e-02 1.02922094e+00 -5.82839727e-01
-4.55392301e-01 6.15130186e-01 -1.06312349e-01 -9.43584517e-02
1.11504912e+00 2.80057043e-01 -9.65185091e-02 1.46460354e-01
-9.16317344e-01 7.25144923e-01 3.50235075e-01 1.32874191e-01
6.15260124e-01 -1.37509286e+00 -5.72422206e-01 4.42910850e-01
-1.19195446e-01 5.33835530e-01 4.27859306e-01 1.36300921e+00
-8.14260066e-01 3.43171775e-01 3.78235914e-02 -7.03878105e-01
-9.82221723e-01 7.51740098e-01 7.52695203e-01 -5.54261863e-01
-8.98895502e-01 7.97754884e-01 8.07499528e-01 -3.71581465e-01
1.88098535e-01 -1.08539999e+00 2.58701742e-02 -4.11616266e-01
4.95032966e-01 -6.74744099e-02 -8.07268694e-02 -2.92003155e-01
-4.50662673e-01 4.99422103e-01 -1.82127550e-01 6.70631647e-01
1.06858420e+00 -2.77393907e-01 -3.08806330e-01 -1.21570058e-01
1.11650693e+00 -1.90398972e-02 -1.42502236e+00 -4.68477368e-01
-1.61613047e-01 -2.44522825e-01 2.28986576e-01 -9.77198184e-01
-1.53420448e+00 6.93211913e-01 6.61308646e-01 8.09030458e-02
1.40001380e+00 2.31097508e-02 1.01284659e+00 -6.94514289e-02
4.03921166e-03 -6.99451268e-01 1.57640219e-01 3.21401477e-01
7.24122465e-01 -1.53945315e+00 1.47784039e-01 -5.97472847e-01
-4.48579311e-01 1.18773293e+00 7.83074081e-01 -3.05633545e-01
7.21306682e-01 3.97326857e-01 2.04625726e-01 -8.15144032e-02
-3.75925124e-01 -1.47339210e-01 3.45352054e-01 6.20545857e-02
5.05290508e-01 1.84278637e-01 -4.62660134e-01 7.73906767e-01
6.71373725e-01 3.64231169e-02 2.27376565e-01 8.73001337e-01
-4.99701023e-01 -8.17370296e-01 -2.22822696e-01 3.14715892e-01
-8.20882976e-01 -3.52888435e-01 5.94097376e-02 8.96984279e-01
1.54591307e-01 4.27853048e-01 -3.21788751e-02 5.69440126e-02
5.07316506e-03 -3.07203054e-01 3.38405848e-01 -5.91091514e-01
-1.03989983e+00 5.88681757e-01 -5.36662161e-01 -5.37368953e-01
-1.87266722e-01 -4.03696597e-01 -1.63545096e+00 2.82299519e-01
-5.78938603e-01 5.12895770e-02 4.71055299e-01 7.25737751e-01
3.60917568e-01 8.75658870e-01 6.50007486e-01 -7.10313857e-01
-5.90619564e-01 -9.67605770e-01 -3.82923365e-01 3.87564242e-01
4.29045647e-01 -5.34438848e-01 -1.86902419e-01 -1.51775926e-01] | [14.594420433044434, -2.550635576248169] |
0c42a7e6-c897-4152-928b-3358b2c5cf3e | msc-a-dataset-for-macro-management-in | 1710.03131 | null | https://arxiv.org/abs/1710.03131v3 | https://arxiv.org/pdf/1710.03131v3.pdf | MSC: A Dataset for Macro-Management in StarCraft II | Macro-management is an important problem in StarCraft, which has been studied for a long time. Various datasets together with assorted methods have been proposed in the last few years. But these datasets have some defects for boosting the academic and industrial research: 1) There're neither standard preprocessing, parsing and feature extraction procedures nor predefined training, validation and test set in some datasets. 2) Some datasets are only specified for certain tasks in macro-management. 3) Some datasets are either too small or don't have enough labeled data for modern machine learning algorithms such as deep neural networks. So most previous methods are trained with various features, evaluated on different test sets from the same or different datasets, making it difficult to be compared directly. To boost the research of macro-management in StarCraft, we release a new dataset MSC based on the platform SC2LE. MSC consists of well-designed feature vectors, pre-defined high-level actions and final result of each match. We also split MSC into training, validation and test set for the convenience of evaluation and comparison. Besides the dataset, we propose a baseline model and present initial baseline results for global state evaluation and build order prediction, which are two of the key tasks in macro-management. Various downstream tasks and analyses of the dataset are also described for the sake of research on macro-management in StarCraft II. Homepage: https://github.com/wuhuikai/MSC. | ['Kaiqi Huang', 'Junge Zhang', 'Yanqi Zong', 'Huikai Wu'] | 2017-10-09 | null | null | null | null | ['real-time-strategy-games'] | ['playing-games'] | [-2.65391320e-01 -4.64095652e-01 -6.65073276e-01 -4.77643728e-01
-2.22198665e-01 -7.17421949e-01 5.75027287e-01 1.72898889e-01
-2.79647589e-01 5.87027609e-01 3.24642807e-01 -1.47311345e-01
-2.10452318e-01 -9.55193818e-01 -4.60618138e-01 -5.14041901e-01
-4.81609553e-02 6.76219642e-01 6.22432649e-01 -8.20169568e-01
2.70562798e-01 3.51656020e-01 -2.03165317e+00 3.98455143e-01
4.57121462e-01 1.22688878e+00 1.96574777e-01 5.18387675e-01
-4.20124494e-02 7.84090102e-01 -9.00644720e-01 -3.40671033e-01
4.00170863e-01 -2.63609171e-01 -7.60094643e-01 -1.38479754e-01
9.21041146e-02 -7.38218948e-02 -4.26347673e-01 7.92549312e-01
7.98113167e-01 1.41530901e-01 1.82528049e-01 -1.47782063e+00
1.39491409e-01 8.49190831e-01 -3.18127014e-02 2.56572694e-01
1.81354195e-01 3.80747914e-01 8.89074624e-01 -2.46944755e-01
5.29719114e-01 7.95906663e-01 5.95041037e-01 3.87357533e-01
-6.17945910e-01 -7.30463505e-01 -1.14285626e-01 4.77327079e-01
-9.52937007e-01 -4.81168568e-01 4.88234401e-01 -4.21154708e-01
1.02037120e+00 3.43411535e-01 7.65173972e-01 1.14843869e+00
5.69562256e-01 7.92773664e-01 8.41821253e-01 -7.49601275e-02
2.33533457e-01 -4.96301539e-02 4.04577583e-01 6.61409914e-01
8.19548443e-02 3.72130483e-01 -5.48631728e-01 2.26252154e-01
5.95688462e-01 -1.64069533e-01 -1.10000342e-01 -4.63700324e-01
-1.28690922e+00 7.85880148e-01 -9.88137573e-02 3.92879248e-01
-1.08614944e-01 -5.30969389e-02 9.16504741e-01 4.18140173e-01
-1.37323916e-01 2.72367448e-01 -8.34076405e-01 -5.42506218e-01
-6.53946102e-01 5.33457100e-01 9.05739188e-01 1.00225794e+00
6.90802515e-01 1.50912166e-01 -1.98112547e-01 6.20656550e-01
-1.78347766e-01 1.37425438e-01 9.45987582e-01 -7.71889389e-01
7.22273886e-01 7.83605456e-01 6.68323487e-02 -8.58008504e-01
-7.20503151e-01 -5.98553419e-01 -7.80692399e-01 2.03790188e-01
3.04603666e-01 -2.60228693e-01 -6.82013571e-01 1.63982737e+00
3.81585509e-01 8.06842372e-02 2.12414950e-01 1.03798318e+00
1.03887832e+00 5.88808179e-01 -3.59945178e-01 -1.15713052e-01
1.21713972e+00 -1.16253924e+00 -8.25223565e-01 -1.96283579e-01
8.18330407e-01 -8.11419547e-01 1.08037245e+00 6.33965611e-01
-1.00141501e+00 -7.31688559e-01 -1.18478870e+00 2.90431350e-01
-6.35942519e-01 2.54031032e-01 8.68946850e-01 6.95528507e-01
-7.70399570e-01 8.74966204e-01 -9.33595777e-01 -2.30805874e-01
-5.91249391e-02 5.35898924e-01 -4.68608379e-01 5.17877579e-01
-1.46815336e+00 1.07066095e+00 7.83969402e-01 -9.87046286e-02
-1.17704415e+00 -2.45770365e-01 -8.09424818e-01 -2.41233259e-01
5.37279725e-01 -4.25535053e-01 1.49175334e+00 -6.53307140e-01
-1.70419979e+00 7.26661325e-01 3.92653018e-01 -4.73051131e-01
1.97772756e-01 -4.53631550e-01 -6.74484849e-01 -4.33997244e-01
5.12125939e-02 9.72941518e-02 5.98001003e-01 -7.70881116e-01
-1.13270950e+00 -3.99675399e-01 4.37918961e-01 1.50212377e-01
-2.47176871e-01 6.96327165e-02 -8.11021864e-01 -6.27933860e-01
-2.11989939e-01 -8.97716522e-01 -3.06542635e-01 -1.04846656e+00
-4.10437047e-01 -6.94993138e-02 5.17082810e-01 -5.56980431e-01
1.81589746e+00 -1.81186795e+00 1.57282248e-01 -1.87375501e-01
-1.86554179e-01 6.33060813e-01 -6.69903383e-02 8.08486164e-01
-1.32614464e-01 -1.43714413e-01 7.72579983e-02 -9.90412831e-02
1.19508155e-01 4.30801958e-01 -3.19446295e-01 3.71603549e-01
-9.29570124e-02 4.40462857e-01 -8.18761230e-01 -4.81705427e-01
5.44187605e-01 -1.13551646e-01 -4.96904463e-01 2.63524085e-01
-3.12754571e-01 4.40894008e-01 -3.85349751e-01 7.47281075e-01
4.38965648e-01 1.91184402e-01 2.11412102e-01 -5.65676808e-01
-2.66949147e-01 4.69173521e-01 -1.44310141e+00 1.93144274e+00
-3.99234563e-01 1.93589851e-01 -2.59505749e-01 -1.25620770e+00
8.14620256e-01 1.83382452e-01 6.27453685e-01 -7.27150321e-01
4.54260200e-01 1.23292282e-01 1.08601458e-01 -6.98482871e-01
7.52902329e-01 9.32297707e-02 -6.13435149e-01 1.87265664e-01
2.36875847e-01 -9.03461650e-02 7.14321673e-01 3.27040330e-02
1.25367928e+00 3.11583251e-01 1.27534956e-01 -1.11468434e-01
5.04536033e-01 3.52253050e-01 1.19915795e+00 3.90567124e-01
-2.57687449e-01 3.82988781e-01 4.55754727e-01 -6.50889635e-01
-7.92516589e-01 -6.58953667e-01 -4.27019298e-02 1.21550548e+00
3.59773874e-01 -1.10392427e+00 -6.38159394e-01 -8.88567030e-01
-9.90709364e-02 6.29872084e-01 -3.95035803e-01 -3.33845735e-01
-7.80381680e-01 -8.03105414e-01 7.17453957e-01 6.28509879e-01
6.35120392e-01 -1.22109282e+00 -1.04780626e+00 3.03765714e-01
-1.31876647e-01 -1.07865465e+00 -1.59977853e-01 5.55433393e-01
-9.07674074e-01 -1.29554057e+00 2.72373170e-01 -5.85856676e-01
-2.34031286e-02 -7.21934577e-03 1.46340680e+00 -3.01856231e-02
-1.21024571e-01 -4.01612557e-02 -8.01005542e-01 -3.44278783e-01
-4.00207847e-01 3.82033616e-01 4.16727006e-01 -2.31892273e-01
1.55938342e-01 -4.76673722e-01 -3.62769157e-01 7.12098181e-01
-8.69958222e-01 6.80419877e-02 6.22829080e-01 7.92473972e-01
3.18061054e-01 2.69462168e-01 2.71858841e-01 -5.76041281e-01
4.71645385e-01 -4.13276702e-01 -7.87741423e-01 6.88146427e-02
-5.67711234e-01 2.02686816e-01 5.79065740e-01 -2.72380173e-01
-5.89309990e-01 -8.31005871e-02 -3.40078712e-01 -5.16904712e-01
-2.97559530e-01 5.79339862e-01 -5.48778892e-01 3.66994411e-01
4.87321496e-01 1.70423016e-01 -4.05072197e-02 -7.47412264e-01
1.91214114e-01 6.52856529e-01 4.89187866e-01 -5.52615583e-01
5.71721077e-01 5.85919209e-02 -9.31554064e-02 -2.76633561e-01
-8.32937539e-01 -4.36371773e-01 -6.53313577e-01 -1.36472359e-01
6.65692329e-01 -5.62905431e-01 -7.51015067e-01 6.45033956e-01
-8.11927438e-01 -5.86683571e-01 -2.53294975e-01 4.85466987e-01
-5.96148491e-01 2.67267257e-01 -5.91221154e-01 -5.26457846e-01
-4.43788767e-01 -1.46490681e+00 8.36488664e-01 2.64599174e-01
2.26432588e-02 -7.11160183e-01 2.36571237e-01 4.45684433e-01
3.05283278e-01 3.07714999e-01 6.42503262e-01 -7.55676448e-01
-2.19239205e-01 -3.05050164e-01 4.89979178e-01 5.32913148e-01
2.35331625e-01 1.96890622e-01 -4.28809494e-01 -2.83632964e-01
-1.41590521e-01 -4.85069841e-01 6.65558755e-01 1.00359187e-01
1.09528410e+00 -8.48638937e-02 -3.62519264e-01 6.27563775e-01
1.26566648e+00 4.48260874e-01 7.05770314e-01 9.04410958e-01
3.99331778e-01 6.28602326e-01 1.36608922e+00 6.59616053e-01
3.27621162e-01 1.00468171e+00 8.24336886e-01 3.84853244e-01
1.09541416e-01 -1.36226803e-01 7.47753620e-01 1.09576666e+00
-2.56808788e-01 -1.80943772e-01 -9.88858223e-01 4.94485736e-01
-2.05237365e+00 -8.96202445e-01 -1.60547599e-01 2.29246044e+00
7.82197177e-01 3.11642349e-01 3.26718271e-01 5.67952394e-01
7.37293005e-01 2.31567234e-01 -3.89166296e-01 -3.97004306e-01
-7.73024559e-02 4.01166022e-01 5.21647930e-01 4.76330370e-02
-1.46670711e+00 1.25598550e+00 5.77446890e+00 1.09099722e+00
-1.22364569e+00 8.65966901e-02 1.95555463e-01 -2.81165272e-01
2.49008432e-01 2.33207524e-01 -1.22406983e+00 6.26439929e-01
1.10834956e+00 8.25124234e-02 5.25354594e-02 9.16695416e-01
1.72566071e-01 -3.17952521e-02 -9.87083912e-01 1.09678125e+00
-1.57976344e-01 -1.46810579e+00 -8.66988301e-02 -1.34316698e-01
4.09895867e-01 1.48205847e-01 -1.78064167e-01 7.39155173e-01
5.10961056e-01 -7.43233979e-01 9.40787494e-01 2.81730562e-01
5.72591305e-01 -8.97755682e-01 1.04243433e+00 5.35118759e-01
-1.18186641e+00 -3.34594727e-01 -4.07721281e-01 -1.20205708e-01
2.85440296e-01 3.88328910e-01 -3.01813304e-01 1.03173828e+00
9.54911292e-01 7.72084594e-01 -5.58223009e-01 1.06264842e+00
-2.68325359e-01 6.55689299e-01 -1.75313547e-01 1.00965612e-01
2.81306416e-01 -2.52783746e-01 4.29127485e-01 8.82822156e-01
1.97831362e-01 -1.26287326e-01 2.18427181e-01 7.65385851e-02
3.44342798e-01 1.49376273e-01 -3.72726202e-01 -1.31323293e-01
3.19308043e-01 1.47035205e+00 -4.07379121e-01 -3.55989456e-01
-2.63622999e-01 5.37389874e-01 -1.64252929e-02 -2.88184822e-01
-1.16891420e+00 -3.26599211e-01 8.10668230e-01 2.92148173e-01
7.47634098e-02 -2.85207093e-01 -2.86913309e-02 -1.33486211e+00
-2.20670462e-01 -1.35253000e+00 6.41953468e-01 -3.22285444e-01
-9.10753787e-01 6.75764382e-01 1.71002388e-01 -1.56043565e+00
-3.51286858e-01 -6.72152519e-01 -6.25606894e-01 2.66938984e-01
-1.00490904e+00 -9.55808520e-01 -4.22373950e-01 6.36236191e-01
6.90616488e-01 -7.08060980e-01 8.06636035e-01 6.03662610e-01
-1.22976458e+00 5.25394619e-01 4.79091741e-02 2.47005075e-01
8.09952378e-01 -1.06804788e+00 4.18363035e-01 7.67596602e-01
7.07818791e-02 4.92854267e-01 8.03725958e-01 -5.91546595e-01
-1.77991688e+00 -1.05235994e+00 4.45376515e-01 -4.82648104e-01
8.18275094e-01 -2.02855363e-01 -3.56041640e-01 7.10991085e-01
1.97757319e-01 -6.33073300e-02 6.63180590e-01 1.21827208e-01
1.56221792e-01 -4.19590324e-01 -8.33992600e-01 3.51389527e-01
1.23089480e+00 1.07165076e-01 -6.50592268e-01 2.39476427e-01
3.92814279e-01 -5.95243275e-01 -1.11652327e+00 8.29885960e-01
4.25254375e-01 -1.43587315e+00 7.68532097e-01 -9.62952912e-01
4.36189532e-01 -3.80869001e-01 -4.58613038e-01 -1.16659796e+00
-1.90517381e-01 -6.24443948e-01 -3.11616182e-01 1.46588814e+00
1.65840551e-01 -4.79236782e-01 9.06476319e-01 2.30587959e-01
-6.45113349e-01 -8.30604315e-01 -6.15567148e-01 -8.75752687e-01
-1.30445316e-01 -6.08483970e-01 9.13577914e-01 8.45086038e-01
-2.10032966e-02 5.41719258e-01 -4.32204694e-01 4.71995324e-02
1.08343214e-01 5.28796017e-01 1.30363214e+00 -9.95266795e-01
-4.23506021e-01 -2.84955978e-01 -6.75934494e-01 -7.77337909e-01
-3.90994251e-02 -7.55526245e-01 -1.91386431e-01 -1.50030458e+00
-1.20392233e-01 -7.32728124e-01 -3.57658505e-01 6.94051147e-01
2.60122061e-01 -7.16287047e-02 1.47310674e-01 3.05662632e-01
-5.57645559e-01 5.51084518e-01 1.04931867e+00 -7.33909234e-02
-1.59779817e-01 2.94089258e-01 -3.43827218e-01 6.94787741e-01
1.15384912e+00 -2.08566770e-01 -4.35740024e-01 -3.90699148e-01
1.81170806e-01 2.05080196e-01 5.44111282e-02 -1.37626886e+00
8.88134390e-02 -4.29344207e-01 9.35131982e-02 -9.44708288e-01
3.79418254e-01 -6.47476375e-01 3.03384095e-01 5.51788568e-01
1.03281409e-01 5.25014699e-01 1.07972205e-01 8.71110931e-02
-4.89290476e-01 -3.43558252e-01 6.50622785e-01 -2.03535318e-01
-1.37357175e+00 5.12450695e-01 -1.61771968e-01 -1.73845869e-02
1.21493042e+00 -1.20648928e-01 -3.33686709e-01 -1.39973968e-01
-7.02092886e-01 1.87498108e-01 5.38795054e-01 7.38495529e-01
2.95202821e-01 -1.26504660e+00 -4.45834488e-01 9.54700261e-02
2.80119460e-02 -1.04796169e-02 1.10752158e-01 8.86761248e-01
-7.14714646e-01 4.00937855e-01 -6.66163921e-01 -4.31935459e-01
-1.22221017e+00 7.20175564e-01 3.24977428e-01 -5.92040062e-01
-4.67138201e-01 5.16858220e-01 -2.05116600e-01 -5.41488826e-01
1.08021945e-01 -4.33595389e-01 -6.29948139e-01 1.69552177e-01
2.79543102e-01 3.36260021e-01 2.54429281e-01 -8.74472380e-01
-4.58200336e-01 4.44014996e-01 1.67592317e-01 5.56417704e-02
1.50399864e+00 2.13203609e-01 -2.49166191e-02 4.39489782e-01
7.00230300e-01 -1.19281568e-01 -7.62927949e-01 8.81437138e-02
1.41869456e-01 -3.73587698e-01 2.31631175e-02 -6.91585541e-01
-1.48982418e+00 8.03541958e-01 7.42922187e-01 7.20720068e-02
1.10776329e+00 -2.37695277e-01 9.80521798e-01 3.08149010e-01
7.42948830e-01 -1.46607554e+00 3.60754062e-03 8.24666142e-01
8.07883441e-01 -1.02463424e+00 -1.39845133e-01 -1.83488607e-01
-7.38060832e-01 1.01595020e+00 7.57797420e-01 -2.96862070e-02
5.07351995e-01 6.00281656e-01 9.95641798e-02 -1.10081874e-01
-9.13097501e-01 -5.54542780e-01 -1.43954262e-01 4.38983381e-01
2.96333730e-01 -3.06706708e-02 -5.80723047e-01 1.17810118e+00
-6.45422101e-01 -1.21255992e-02 3.21728170e-01 1.24311554e+00
-4.63375837e-01 -1.73458183e+00 -2.63232499e-01 4.68983889e-01
-5.46899140e-01 2.01609299e-01 -1.66481584e-01 9.46267545e-01
6.20683610e-01 1.02761340e+00 -1.44344702e-01 -1.16755462e+00
7.43530452e-01 -1.40715525e-01 3.95600945e-01 -5.74023426e-01
-1.09196484e+00 -3.01203281e-01 6.07890189e-01 -8.77174139e-01
-1.98459432e-01 -4.71703649e-01 -1.18371546e+00 -4.98031884e-01
-3.77087831e-01 3.67378503e-01 6.78312600e-01 7.71655858e-01
2.91001797e-01 5.79890251e-01 5.42143524e-01 -8.56600285e-01
-7.46109605e-01 -1.04016805e+00 -4.98598397e-01 5.27136981e-01
-2.52186418e-01 -9.94557023e-01 -1.01845585e-01 -9.18990448e-02] | [4.086453437805176, 1.5883162021636963] |
49eb609b-3857-42ea-bf91-b842d5cf7524 | temporal-relational-modeling-with-self | 2012.07508 | null | https://arxiv.org/abs/2012.07508v1 | https://arxiv.org/pdf/2012.07508v1.pdf | Temporal Relational Modeling with Self-Supervision for Action Segmentation | Temporal relational modeling in video is essential for human action understanding, such as action recognition and action segmentation. Although Graph Convolution Networks (GCNs) have shown promising advantages in relation reasoning on many tasks, it is still a challenge to apply graph convolution networks on long video sequences effectively. The main reason is that large number of nodes (i.e., video frames) makes GCNs hard to capture and model temporal relations in videos. To tackle this problem, in this paper, we introduce an effective GCN module, Dilated Temporal Graph Reasoning Module (DTGRM), designed to model temporal relations and dependencies between video frames at various time spans. In particular, we capture and model temporal relations via constructing multi-level dilated temporal graphs where the nodes represent frames from different moments in video. Moreover, to enhance temporal reasoning ability of the proposed model, an auxiliary self-supervised task is proposed to encourage the dilated temporal graph reasoning module to find and correct wrong temporal relations in videos. Our DTGRM model outperforms state-of-the-art action segmentation models on three challenging datasets: 50Salads, Georgia Tech Egocentric Activities (GTEA), and the Breakfast dataset. The code is available at https://github.com/redwang/DTGRM. | ['Dejing Dou', 'Xingjian Li', 'Di Hu', 'Dong Wang'] | 2020-12-14 | null | null | null | null | ['action-understanding'] | ['computer-vision'] | [ 1.01047894e-02 4.35526818e-02 -4.21611369e-01 -3.00241530e-01
1.16766594e-01 -2.94447243e-01 4.22656178e-01 -2.04069108e-01
-2.09941268e-01 3.25269639e-01 2.40013063e-01 -2.78180003e-01
-3.19416434e-01 -6.99365735e-01 -6.77455008e-01 -4.78415936e-01
-2.55943418e-01 1.51190490e-01 6.41607642e-01 -1.06229633e-01
-8.87098908e-02 4.08541530e-01 -1.16160917e+00 4.14812237e-01
8.35105181e-01 9.08278525e-01 6.13068044e-02 6.82638526e-01
-2.92314701e-02 1.57642019e+00 -1.54320434e-01 -5.06399810e-01
4.39836234e-02 -7.25371659e-01 -1.13097918e+00 2.39953578e-01
2.08234087e-01 -5.40753663e-01 -1.05085373e+00 1.03582525e+00
3.56042199e-02 6.27570391e-01 2.62874454e-01 -1.72250366e+00
-5.79365194e-01 8.53590727e-01 -5.51944315e-01 6.79809451e-01
3.03009152e-01 2.34101236e-01 9.17425156e-01 -2.47563794e-01
7.54121661e-01 1.39677191e+00 4.89395112e-01 5.52512586e-01
-5.80516040e-01 -6.25845432e-01 6.75404608e-01 1.00632930e+00
-1.05261326e+00 -1.71153784e-01 8.30116868e-01 -3.58632505e-01
1.03244638e+00 6.51310533e-02 1.06609797e+00 1.09229827e+00
2.21737981e-01 1.15242064e+00 5.44414759e-01 1.50029853e-01
-1.45166218e-02 -6.89649940e-01 1.37655675e-01 8.11638355e-01
-3.67175400e-01 -2.77897328e-01 -5.19413054e-01 3.26628774e-01
1.02780879e+00 3.44778568e-01 -2.52560496e-01 -2.83504188e-01
-1.28865540e+00 5.68250835e-01 6.03271246e-01 3.52303386e-01
-4.38527912e-01 5.36146164e-01 6.28134608e-01 1.19886294e-01
3.98695022e-01 -2.36034453e-01 -3.98087025e-01 -3.57540905e-01
-4.57382888e-01 1.02867961e-01 5.06805837e-01 9.63832796e-01
2.66098231e-01 7.49711835e-05 -2.38606900e-01 7.55640864e-01
2.49422252e-01 3.03445347e-02 3.23594034e-01 -1.22206068e+00
6.64621115e-01 9.51613307e-01 -2.87899643e-01 -1.36607587e+00
-5.25476515e-01 -1.52787799e-02 -9.61172819e-01 -3.76249284e-01
5.77857256e-01 2.26150732e-02 -8.25899899e-01 1.59837484e+00
4.95710343e-01 7.43522704e-01 3.09432652e-02 9.77614880e-01
1.04500806e+00 6.50037944e-01 3.30748260e-01 -2.43352965e-01
1.31596839e+00 -1.29396057e+00 -9.38177347e-01 -1.68038592e-01
8.28175843e-01 -2.94865996e-01 6.90175593e-01 1.23056151e-01
-1.03948987e+00 -6.45729065e-01 -5.53542256e-01 -2.16677681e-01
-2.64817089e-01 4.74450774e-02 8.57016087e-01 -1.70896173e-01
-8.74227107e-01 6.25573218e-01 -1.25573325e+00 -6.09112918e-01
6.91175938e-01 2.56994992e-01 -3.25043678e-01 -2.13653207e-01
-1.37048876e+00 5.46121478e-01 6.40333116e-01 6.06289804e-01
-8.72565746e-01 -1.99106738e-01 -1.02168512e+00 -6.85459450e-02
8.33307981e-01 -5.06678224e-01 1.20325720e+00 -9.25218582e-01
-1.11908913e+00 5.20801723e-01 4.90060337e-02 -7.71545887e-01
7.04112351e-01 -3.04391533e-01 -4.67080981e-01 5.87437570e-01
1.75515877e-03 6.96759641e-01 5.78992248e-01 -5.49507856e-01
-5.38477957e-01 -3.26979995e-01 5.75025558e-01 3.35618198e-01
-1.77463859e-01 2.94951387e-02 -1.00894165e+00 -6.81382358e-01
2.46479154e-01 -1.08227158e+00 -2.31113106e-01 -1.03797548e-01
-3.64024341e-01 -5.33035755e-01 1.09090698e+00 -9.37256753e-01
1.30687261e+00 -2.05339098e+00 3.71727675e-01 -1.95640653e-01
2.39464402e-01 4.38279837e-01 -1.41762063e-01 2.80703664e-01
-2.20847905e-01 -6.22390956e-02 -7.98902065e-02 -4.03720289e-02
-1.54697567e-01 7.09176660e-01 -2.84089940e-03 3.85830224e-01
7.03812316e-02 1.14469981e+00 -1.09896612e+00 -7.42862105e-01
4.95712578e-01 5.22348344e-01 -3.76136601e-01 8.76577664e-03
-5.13931692e-01 7.43235528e-01 -7.42197573e-01 6.06520891e-01
3.62776518e-01 -4.58537519e-01 2.56676674e-01 -4.17458087e-01
1.15913443e-01 1.47708282e-01 -1.05143547e+00 1.92575383e+00
2.46118587e-02 6.18700445e-01 -4.52672809e-01 -1.15403283e+00
5.16226768e-01 4.19527024e-01 9.22579229e-01 -8.84543538e-01
3.78836483e-01 -3.37410569e-01 1.54064506e-01 -9.13418174e-01
2.64998615e-01 3.78139585e-01 1.87868327e-01 2.64378488e-01
1.68041408e-01 3.42364997e-01 7.20784366e-01 5.50063491e-01
1.21574259e+00 5.87092340e-01 6.89523667e-02 2.30742052e-01
7.03847885e-01 3.07485135e-03 9.33754325e-01 3.59179497e-01
-6.01104975e-01 3.03421795e-01 9.65013266e-01 -6.26019359e-01
-6.41378224e-01 -6.21241271e-01 4.74522680e-01 9.32527900e-01
4.76761013e-01 -6.84406161e-01 -7.84567893e-01 -9.45763826e-01
-2.96539962e-01 4.42594707e-01 -7.09241033e-01 -2.56610870e-01
-8.88440788e-01 -4.79733288e-01 4.58373189e-01 9.71839249e-01
1.09972954e+00 -1.36557698e+00 -5.04619777e-01 2.14127079e-01
-6.46114886e-01 -1.68349421e+00 -5.38355529e-01 -3.68666887e-01
-1.01160252e+00 -1.52356374e+00 -3.80954206e-01 -5.96716523e-01
4.91905957e-01 4.57860202e-01 1.02107000e+00 2.32972175e-01
-1.06236435e-01 5.26435375e-01 -6.52430415e-01 8.20599794e-02
-7.74476156e-02 -1.37939334e-01 -2.19431072e-01 1.34473041e-01
4.04393613e-01 -6.62995160e-01 -7.36156583e-01 5.38816571e-01
-9.09244895e-01 4.63675737e-01 2.47645542e-01 5.36767960e-01
6.70127392e-01 4.03324068e-01 1.33305773e-01 -8.23638320e-01
2.41072625e-01 -4.08832997e-01 -4.62063581e-01 3.34728330e-01
-9.03291777e-02 -3.35620701e-01 5.13296902e-01 -5.29315114e-01
-1.03561354e+00 -1.16092153e-02 -3.20862769e-03 -8.69075894e-01
-2.76009627e-02 5.08700848e-01 -1.15333192e-01 2.24866331e-01
2.93174148e-01 1.46524042e-01 -1.10359259e-01 -1.54213458e-01
3.94465923e-01 9.01335403e-02 5.92508316e-01 -3.92944425e-01
4.37363386e-01 5.14198422e-01 8.93878192e-02 -5.55099487e-01
-9.01580155e-01 -5.20208180e-01 -8.66253912e-01 -6.49849534e-01
1.44589984e+00 -9.18883920e-01 -8.99549603e-01 7.33629107e-01
-1.09912932e+00 -8.46449912e-01 -8.18613619e-02 6.20545030e-01
-6.41046941e-01 5.17797053e-01 -9.56341267e-01 -6.09627903e-01
-1.06819592e-01 -1.01731288e+00 7.39661992e-01 4.54738259e-01
-5.96060008e-02 -1.18592322e+00 -3.39911550e-01 8.00467253e-01
-1.80217832e-01 4.66763645e-01 5.83075285e-01 -2.25984141e-01
-8.08718145e-01 7.86468238e-02 -3.59311849e-01 3.61615837e-01
1.53461397e-01 1.34312138e-01 -3.59018058e-01 1.22931458e-01
-2.07388073e-01 -2.41722673e-01 8.98018479e-01 5.26372135e-01
1.37795997e+00 -2.24810630e-01 -2.64417678e-01 6.25900984e-01
8.42736602e-01 6.04720414e-01 9.44908917e-01 2.32765228e-01
1.30005801e+00 4.88818020e-01 9.32374656e-01 3.71761262e-01
7.10069835e-01 4.75507498e-01 5.15307009e-01 8.27134922e-02
-1.80601403e-01 -2.48211637e-01 2.95976192e-01 7.79826224e-01
-5.00697076e-01 -4.20752943e-01 -9.11886215e-01 5.00898123e-01
-2.57803392e+00 -1.16640174e+00 -4.63450521e-01 1.50911605e+00
4.12462920e-01 1.79190502e-01 1.93403453e-01 -5.53540960e-02
6.96124613e-01 4.36796039e-01 -7.62695193e-01 4.91470769e-02
4.10841964e-02 -2.32120991e-01 2.34477088e-01 2.33345151e-01
-1.22168350e+00 1.22502983e+00 4.31360674e+00 6.42526329e-01
-7.61006773e-01 1.21749461e-01 5.76971889e-01 -7.36507995e-04
1.98432401e-01 8.71984884e-02 -3.36990774e-01 3.58778417e-01
6.22954488e-01 4.93166484e-02 4.91164565e-01 6.46178603e-01
4.50449198e-01 -1.98727727e-01 -9.75589037e-01 1.09631729e+00
-2.29411572e-02 -1.16054118e+00 4.87679709e-03 -2.47437701e-01
5.72545528e-01 -1.24588728e-01 -2.68478930e-01 3.61200839e-01
3.47350329e-01 -8.04775774e-01 4.16817188e-01 6.45647824e-01
2.23207563e-01 -6.44269764e-01 5.22559643e-01 2.35216901e-01
-1.65287220e+00 -1.46249115e-01 -1.17567286e-01 -1.34148791e-01
2.51780182e-01 1.12047911e-01 -4.28993642e-01 7.99430311e-01
1.00459206e+00 1.47191894e+00 -5.95321953e-01 7.99764633e-01
-5.42365491e-01 4.91202712e-01 -3.79342549e-02 3.17636639e-01
5.09262085e-01 -5.65934956e-01 3.39970320e-01 8.03614140e-01
8.87882337e-02 6.24381781e-01 2.27630585e-01 5.65413713e-01
3.15155871e-02 -2.41047800e-01 -4.39612567e-01 -5.01271069e-01
3.32725197e-02 1.24255884e+00 -1.05341792e+00 -3.87521684e-01
-5.83904088e-01 1.03505814e+00 3.06185514e-01 5.87116063e-01
-1.28325486e+00 2.17770502e-01 6.44631267e-01 -1.01750225e-01
2.38842353e-01 -5.05950630e-01 2.29339063e-01 -1.34400725e+00
5.53591587e-02 -8.21481824e-01 1.02639902e+00 -1.09266365e+00
-9.01614845e-01 4.32609022e-01 2.22452462e-01 -1.33030117e+00
-1.86461300e-01 -3.13148558e-01 -5.40566504e-01 3.15307796e-01
-1.11838102e+00 -1.21873474e+00 -7.30474830e-01 1.06347644e+00
8.02719593e-01 1.65614665e-01 1.61908522e-01 4.58376676e-01
-9.33712423e-01 1.92605913e-01 -6.46727681e-01 6.48771167e-01
2.32738167e-01 -1.02522349e+00 4.83776569e-01 1.01336336e+00
2.31115520e-01 3.62506986e-01 3.91525030e-01 -8.92132342e-01
-1.30853593e+00 -1.40117919e+00 5.31074405e-01 -1.92062408e-01
9.27702785e-01 -2.49024648e-02 -1.03130531e+00 1.11065412e+00
1.49492115e-01 3.63765150e-01 3.06663722e-01 -1.54018134e-01
-1.11148730e-01 3.20794247e-02 -6.45170331e-01 8.80672038e-01
1.65142119e+00 -4.88909572e-01 -3.20672870e-01 7.27320194e-01
8.12788904e-01 -7.57274330e-01 -8.79553676e-01 4.56701905e-01
3.87791842e-01 -9.44017410e-01 9.22939003e-01 -6.77667737e-01
6.46504700e-01 -4.51966405e-01 7.35342577e-02 -9.24785674e-01
-1.86578944e-01 -5.71732521e-01 -5.35692275e-01 1.15963840e+00
-1.75379515e-01 -5.04743099e-01 9.06360745e-01 6.05532885e-01
-2.26895168e-01 -9.15358901e-01 -6.70166969e-01 -7.24416733e-01
-5.40528417e-01 -5.91205478e-01 3.64772826e-01 1.08124435e+00
-6.55247942e-02 2.54371822e-01 -5.21909297e-01 1.82155401e-01
3.85900766e-01 -2.04064026e-02 6.70253575e-01 -7.88087845e-01
-2.43972287e-01 -4.04118001e-01 -7.06550241e-01 -1.11064219e+00
3.39176774e-01 -6.87245548e-01 -2.28097275e-01 -1.93966043e+00
1.02729283e-01 -1.40708491e-01 -3.27695221e-01 7.52798200e-01
-8.59497488e-02 8.09204578e-02 2.88749874e-01 5.56248166e-02
-1.06663537e+00 6.51555598e-01 1.71635485e+00 -2.94504404e-01
-2.12840676e-01 1.14407297e-02 -2.02376351e-01 9.03831363e-01
8.10837150e-01 -2.32505962e-01 -9.79478717e-01 -4.71703738e-01
1.10896252e-01 5.93578160e-01 7.19629228e-01 -1.14325666e+00
4.27610755e-01 -3.32941085e-01 2.86432952e-01 -8.34616303e-01
3.39001626e-01 -7.25069225e-01 4.51076746e-01 5.38641572e-01
-3.07262659e-01 1.87360734e-01 6.71632290e-02 6.77046299e-01
-4.04052079e-01 1.73222005e-01 4.15295959e-01 -3.02141905e-01
-1.29115760e+00 9.03612435e-01 -1.11819901e-01 8.19885582e-02
1.19195867e+00 -1.76893383e-01 -4.73987818e-01 -4.33081508e-01
-1.07511663e+00 7.50835061e-01 9.81461182e-02 6.56851172e-01
6.81713402e-01 -1.42973840e+00 -3.04501414e-01 -3.39674622e-01
2.86498154e-03 3.33283931e-01 9.46327269e-01 1.37083089e+00
-5.98523140e-01 1.79000527e-01 -2.77416706e-01 -6.83835804e-01
-1.42885244e+00 6.33675456e-01 5.14160097e-01 -2.52787858e-01
-1.02339506e+00 9.27350104e-01 3.09178859e-01 -4.42328565e-02
2.74649233e-01 -6.58279777e-01 -5.40177286e-01 9.55379084e-02
3.27587098e-01 3.50488931e-01 -3.38118702e-01 -7.43144989e-01
-3.70217055e-01 4.25155282e-01 -3.02791782e-02 3.02599758e-01
1.10903156e+00 -1.43303335e-01 -2.04768509e-01 4.09107149e-01
9.78876770e-01 -7.79254854e-01 -1.63797545e+00 -3.06829840e-01
-2.40654215e-01 -4.03359115e-01 -1.53740734e-01 -3.56574029e-01
-1.59616351e+00 7.72546709e-01 3.01590234e-01 2.19544619e-01
1.30180657e+00 4.53172848e-02 1.12906778e+00 2.87464976e-01
1.98516011e-01 -1.18836391e+00 4.65144753e-01 5.49148798e-01
7.34562278e-01 -1.09751701e+00 1.92938035e-03 -6.63681448e-01
-9.11062062e-01 1.05575919e+00 9.57324266e-01 -2.68665925e-02
4.89757508e-01 -2.21331075e-01 -1.37000248e-01 -4.67112511e-01
-6.74905539e-01 -4.04446542e-01 2.92329282e-01 4.17863041e-01
2.09860906e-01 -2.33497974e-02 -2.69583017e-01 3.50243449e-01
1.66321158e-01 2.44390726e-01 3.14373285e-01 8.87783945e-01
6.15060963e-02 -8.85050058e-01 6.81369603e-02 3.89262885e-01
-3.79560679e-01 1.80641845e-01 -2.89007604e-01 9.61755693e-01
2.93135166e-01 1.04725957e+00 6.82404861e-02 -6.24084830e-01
2.16904357e-01 -1.13390483e-01 4.88702565e-01 -2.13747516e-01
-3.22369277e-01 5.26532382e-02 2.03856394e-01 -1.06882632e+00
-9.08923626e-01 -8.96547735e-01 -1.69479442e+00 -2.96301037e-01
6.34045824e-02 -7.43229315e-02 -2.31258683e-02 1.17836666e+00
1.69746235e-01 1.09747696e+00 1.24645114e-01 -5.20712972e-01
2.53236383e-01 -8.74259710e-01 -5.46110332e-01 6.96089864e-01
-4.99559827e-02 -8.24250817e-01 9.37692076e-02 3.77243996e-01] | [8.536276817321777, 0.6374086737632751] |
b72f6813-d83c-4a0e-b922-73a8a612e18d | video-story-composition-via-plot-analysis | null | null | http://openaccess.thecvf.com/content_cvpr_2016/html/Choi_Video-Story_Composition_via_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Choi_Video-Story_Composition_via_CVPR_2016_paper.pdf | Video-Story Composition via Plot Analysis | We address the problem of composing a story out of multiple short video clips taken by a person during an activity or experience. Inspired by plot analysis of written stories, our method generates a sequence of video clips ordered in such a way that it reflects plot dynamics and content coherency. That is, given a set of multiple video clips, our method composes a video which we call a video-story. We define metrics on scene dynamics and coherency by dense optical flow features and a patch matching algorithm. Using these metrics, we define an objective function for the video-story. To efficiently search for the best video-story, we introduce a novel Branch-and-Bound algorithm which guarantees the global optimum. We collect the dataset consisting of 23 video sets from the web, resulting in a total of 236 individual video clips. With the acquired dataset, we perform extensive user studies involving 30 human subjects by which the effectiveness of our approach is quantitatively and qualitatively verified. | ['Tae-Hyun Oh', 'Jinsoo Choi', 'In So Kweon'] | 2016-06-01 | null | null | null | cvpr-2016-6 | ['patch-matching'] | ['computer-vision'] | [ 2.17058927e-01 -4.01135236e-01 -1.70561448e-01 -2.02203929e-01
-5.63840806e-01 -7.72031724e-01 4.29796189e-01 1.90441176e-01
6.95777461e-02 3.40014428e-01 6.66548431e-01 2.78637171e-01
-2.28068128e-01 -6.05236828e-01 -7.21402645e-01 -3.75368774e-01
-2.70622522e-01 -1.49924144e-01 2.61996388e-01 1.05352320e-01
5.47195554e-01 2.14077592e-01 -1.60749483e+00 5.70995510e-01
6.18206441e-01 9.88804281e-01 3.89981061e-01 8.60349059e-01
1.43354729e-01 1.18201780e+00 -3.28425825e-01 -3.60973716e-01
3.59510452e-01 -8.87972832e-01 -7.33560622e-01 7.19716012e-01
5.18109262e-01 -3.76481563e-01 -4.92695838e-01 9.00115013e-01
8.41574296e-02 4.17712569e-01 3.47629160e-01 -1.25427794e+00
-2.67385036e-01 5.95130622e-01 -5.95352709e-01 3.52697641e-01
1.05563343e+00 5.93514787e-03 1.25013471e+00 -7.45802283e-01
1.26168180e+00 8.47854912e-01 4.26394910e-01 -2.23194242e-01
-1.21273994e+00 -2.22252220e-01 8.63143131e-02 4.42379117e-01
-1.46505928e+00 -5.05102277e-01 1.01227283e+00 -7.09036887e-01
4.14235920e-01 4.40999418e-01 1.40460587e+00 6.96541071e-01
1.05205417e-01 8.99534881e-01 5.62107146e-01 -1.94774181e-01
3.60558927e-01 -2.84579724e-01 -1.78075239e-01 6.64637625e-01
-1.25617534e-01 -3.84387463e-01 -9.67034280e-01 -2.46363625e-01
7.86182821e-01 1.07579313e-01 -5.30580878e-01 -7.18325198e-01
-1.45130837e+00 6.25967026e-01 6.78937435e-02 2.41730705e-01
-5.66350460e-01 -3.16874385e-02 2.86679953e-01 2.99347162e-01
2.23176792e-01 3.49774152e-01 3.05168778e-01 -4.38124359e-01
-1.40069938e+00 7.01649666e-01 8.78550887e-01 9.65489745e-01
5.38979352e-01 -4.63016182e-01 -3.48390549e-01 6.28945053e-01
4.82543185e-02 2.20091775e-01 6.80260137e-02 -1.28054500e+00
4.28400069e-01 4.77920294e-01 2.44404882e-01 -1.74378133e+00
-1.42157614e-01 -3.59946266e-02 -4.36613739e-01 -2.69407004e-01
3.34218562e-01 1.78205237e-01 -1.53573304e-01 1.54582548e+00
3.20185125e-01 6.96625233e-01 -4.43032265e-01 1.16934311e+00
5.06282687e-01 8.67119551e-01 -4.61441755e-01 -8.05902779e-01
1.21818244e+00 -9.32944834e-01 -7.88055599e-01 2.97884464e-01
1.73344597e-01 -7.54779160e-01 9.51554835e-01 6.80084705e-01
-1.46007895e+00 -4.26897496e-01 -1.26544011e+00 1.73450738e-01
3.12944084e-01 -2.43201539e-01 9.21381712e-02 2.47237280e-01
-9.11919355e-01 6.44553840e-01 -5.33710897e-01 -5.77703714e-01
1.95963636e-01 -1.72101721e-01 -3.47036541e-01 -6.04502484e-02
-6.31645203e-01 3.07163954e-01 1.94126979e-01 -6.19193390e-02
-1.05660701e+00 -7.26243138e-01 -6.33500874e-01 4.00153548e-02
5.86090505e-01 -4.88785565e-01 1.11523294e+00 -1.16697371e+00
-1.42857611e+00 8.78915310e-01 -1.35428935e-01 -3.35282236e-01
5.28239787e-01 -3.22927147e-01 -3.03936571e-01 7.21688092e-01
3.09459455e-02 2.06634998e-01 9.17631805e-01 -1.23734021e+00
-6.94062710e-01 -3.98812033e-02 3.89514655e-01 4.24630702e-01
-5.64120173e-01 1.70938313e-01 -9.26856101e-01 -7.65230238e-01
4.14973386e-02 -8.01256716e-01 9.40596983e-02 -3.63907218e-02
-3.52608174e-01 3.02527308e-01 6.12736523e-01 -7.83241272e-01
1.88930523e+00 -2.36932635e+00 6.35803938e-01 3.20801467e-01
4.84007150e-01 -3.76636744e-01 -2.00811446e-01 6.47093177e-01
2.05885589e-01 -4.26674783e-02 -2.10952207e-01 -5.98985180e-02
-2.74749011e-01 -2.10851938e-01 -2.56166309e-01 6.64544582e-01
-1.11178927e-01 4.65415001e-01 -1.15435803e+00 -7.74515390e-01
1.06933743e-01 1.38723612e-01 -8.64645898e-01 5.22883892e-01
-3.03409010e-01 4.11873966e-01 -4.59934771e-01 5.99399745e-01
5.31685412e-01 -2.83746988e-01 3.59487265e-01 -2.53045201e-01
-3.39161426e-01 -1.14333205e-01 -1.34227252e+00 2.16349220e+00
-2.70981640e-02 9.31983650e-01 -9.02558491e-02 -7.35799491e-01
7.14038670e-01 2.18765765e-01 1.11593664e+00 -5.29251754e-01
2.21913010e-01 -1.32220164e-01 -4.38241541e-01 -9.52327311e-01
7.49174356e-01 2.54107922e-01 -1.04603864e-01 5.58400750e-01
-6.77467138e-02 -1.27887065e-02 6.09028161e-01 4.29682344e-01
1.36948776e+00 2.51821548e-01 4.23767090e-01 -2.17568755e-01
3.63599628e-01 1.01662815e-01 3.43361855e-01 7.04182863e-01
-3.46584767e-01 9.70343292e-01 6.71567917e-01 -4.99321878e-01
-1.26756668e+00 -1.10839605e+00 2.84921736e-01 8.77151787e-01
4.85810816e-01 -1.02425623e+00 -9.64676440e-01 -1.60928592e-01
-2.40923032e-01 1.34408697e-01 -5.30277431e-01 1.74500972e-01
-5.61702073e-01 -7.09218681e-02 9.64164454e-03 8.25334340e-03
4.02611434e-01 -9.90285754e-01 -9.98539090e-01 2.95161754e-01
-6.44349337e-01 -1.21095800e+00 -1.00943887e+00 -7.04972565e-01
-5.87296605e-01 -1.20304918e+00 -6.89177454e-01 -7.44979143e-01
3.90147448e-01 7.63987124e-01 1.19138515e+00 1.04419194e-01
-2.34774083e-01 4.02889401e-01 -7.10641623e-01 3.43672961e-01
-7.90679976e-02 -2.66522974e-01 5.40955812e-02 5.89689612e-01
-2.39126951e-01 -8.40958595e-01 -6.59651458e-01 3.59402448e-01
-9.85190332e-01 3.78242612e-01 9.58118662e-02 4.35807526e-01
7.61895537e-01 1.10994533e-01 -4.04586978e-02 -3.78456473e-01
6.33547425e-01 -6.46799207e-01 -4.95123714e-01 3.83665174e-01
6.71537817e-02 -2.81867117e-01 5.55147171e-01 -5.29235423e-01
-8.01461279e-01 2.46512085e-01 6.34936094e-01 -7.33760059e-01
2.34771729e-01 7.38132298e-01 -1.27424419e-01 6.35218173e-02
5.34994602e-01 9.79604665e-03 -2.39594981e-01 -2.37358108e-01
4.10088599e-01 4.08750862e-01 8.46892357e-01 -4.14972514e-01
6.89527392e-01 5.44680476e-01 -3.55326653e-01 -1.05582917e+00
-6.95425272e-01 -6.83461607e-01 -6.46194041e-01 -1.04621100e+00
8.83982658e-01 -7.38191426e-01 -8.16698074e-01 2.46309787e-01
-1.16312480e+00 1.38911819e-02 -9.36372876e-02 4.81694400e-01
-9.48756397e-01 3.30989480e-01 -4.08909917e-01 -6.44816756e-01
1.62595049e-01 -8.57715070e-01 7.75693893e-01 1.79752275e-01
-5.51870763e-01 -6.69101179e-01 3.85488451e-01 1.96104437e-01
-6.81598186e-02 7.94314802e-01 4.07033712e-01 -6.20897822e-02
-8.98646891e-01 -1.05966628e-01 -7.70657361e-02 -2.30649695e-01
8.71362165e-02 3.55240971e-01 -5.69181919e-01 -2.99392432e-01
2.58636731e-03 -3.47485580e-02 4.86485749e-01 6.08162701e-01
1.07856572e+00 -5.24115562e-01 -9.53320339e-02 7.09473968e-01
1.59630954e+00 4.29631799e-01 6.47397637e-01 3.95636678e-01
7.68246353e-01 5.85153520e-01 7.05399096e-01 9.13675666e-01
2.06548274e-01 8.32510710e-01 2.69817024e-01 4.35808122e-01
1.61087006e-01 -5.76344490e-01 4.60625172e-01 1.09264982e+00
-3.10538113e-01 -5.03084481e-01 -5.88360608e-01 5.73441505e-01
-2.17462087e+00 -1.46092021e+00 1.94449648e-01 2.23552036e+00
4.26084489e-01 1.01014204e-01 6.40649319e-01 1.68073773e-01
8.69843602e-01 5.18579006e-01 -3.65962148e-01 -7.33226910e-02
1.64986216e-02 -3.41284931e-01 7.30955154e-02 3.59781981e-01
-9.90789652e-01 4.65913385e-01 6.55681849e+00 5.48762083e-01
-8.22719455e-01 -1.44383535e-01 3.37959796e-01 -4.55309957e-01
-4.76321280e-01 9.54767987e-02 -5.74145950e-02 5.06518781e-01
6.79832518e-01 -6.78427339e-01 7.63639390e-01 5.47120214e-01
6.54662311e-01 -3.48673165e-01 -1.07720935e+00 1.32921195e+00
4.56864029e-01 -1.80045474e+00 -5.55879809e-02 -8.13318789e-02
8.56215954e-01 -4.91145968e-01 -1.89460009e-01 -3.78019512e-01
-1.20982580e-01 -7.23474860e-01 1.39783752e+00 6.73302710e-01
6.06334329e-01 -6.90983176e-01 9.23920050e-02 1.54199883e-01
-1.50003994e+00 -5.99981882e-02 7.26404861e-02 -1.30996183e-01
5.02299845e-01 4.17978615e-01 -2.63648897e-01 5.72086930e-01
7.00213253e-01 1.19078183e+00 -2.31300250e-01 1.50320935e+00
1.73558310e-01 4.82485384e-01 -7.24449158e-02 -1.01901047e-01
3.50307003e-02 -5.08516610e-01 8.61672759e-01 1.22169530e+00
4.38505560e-01 3.68407130e-01 3.98351371e-01 6.98494375e-01
-1.25479907e-01 3.63987237e-01 -7.57533431e-01 -2.09180817e-01
4.87075269e-01 1.26108265e+00 -1.03568053e+00 -2.35750258e-01
-3.43716502e-01 1.15758276e+00 1.73372433e-01 3.40919495e-01
-1.01344442e+00 -1.20130204e-01 5.02122283e-01 3.20253044e-01
3.11879307e-01 -3.24977785e-01 -1.76182657e-01 -1.29558063e+00
4.01712239e-01 -7.89889932e-01 2.69029438e-01 -9.57674325e-01
-8.94069612e-01 7.37755716e-01 1.58942059e-01 -1.75664937e+00
-9.56112370e-02 6.58298880e-02 -5.23440003e-01 1.43355057e-01
-6.50927663e-01 -5.62344849e-01 -7.32594073e-01 6.65897787e-01
8.46680462e-01 1.02762245e-02 3.37140590e-01 2.92085499e-01
-3.30279976e-01 1.28187627e-01 5.12355790e-02 -2.50447541e-02
4.04586524e-01 -7.15216637e-01 1.96461275e-01 1.06801200e+00
5.15040755e-01 3.70833158e-01 9.67613757e-01 -6.14083290e-01
-1.46128559e+00 -6.77237809e-01 6.79026186e-01 -2.36147523e-01
8.58762264e-01 -3.71507585e-01 -6.30131662e-01 4.32842255e-01
2.62684613e-01 -2.17446476e-01 6.38743520e-01 -2.59888440e-01
-1.84398219e-01 -9.22933966e-02 -8.66003275e-01 6.37664557e-01
1.48274815e+00 -5.60474277e-01 -4.93484110e-01 1.63846374e-01
7.38514245e-01 -5.13943017e-01 -9.01249051e-01 -1.92231964e-02
9.30518866e-01 -1.18639433e+00 7.70962059e-01 -3.94191712e-01
9.03626263e-01 -3.92045796e-01 -2.94624031e-01 -1.15289736e+00
-5.75155497e-01 -1.12252676e+00 -2.98715323e-01 1.06004977e+00
7.97458179e-03 2.00316012e-01 5.84874332e-01 3.05170447e-01
7.15685487e-02 -5.50470293e-01 -5.01930177e-01 -6.48375034e-01
-8.11236799e-01 -4.33094621e-01 4.36899006e-01 8.22083533e-01
3.77366602e-01 -1.83161311e-02 -6.94011271e-01 -1.34736657e-01
6.87052310e-01 3.59856516e-01 7.31694341e-01 -9.27717209e-01
-2.13456869e-01 -3.11659008e-01 -5.40836632e-01 -1.04370427e+00
-1.11251257e-01 -5.02209902e-01 1.44931406e-01 -1.35011876e+00
6.38114154e-01 -4.07534465e-02 -1.38640821e-01 -1.42184585e-01
1.27774090e-01 1.71185255e-01 5.38112938e-01 5.36144376e-01
-9.37336326e-01 2.36939728e-01 1.29078388e+00 -6.56458214e-02
-4.82324898e-01 -2.62574047e-01 -5.20610690e-01 5.75411201e-01
4.06484693e-01 -2.88068175e-01 -5.99521279e-01 -2.96215177e-01
4.46673840e-01 7.41211891e-01 3.05214882e-01 -1.07045245e+00
1.90887034e-01 -5.40516198e-01 1.07968494e-01 -5.80081105e-01
1.56479806e-01 -6.62247479e-01 6.52219772e-01 1.70162737e-01
-5.57701051e-01 2.28840262e-01 -1.68519601e-01 7.02503145e-01
-3.65591317e-01 -8.80820863e-03 5.28731525e-01 1.96555611e-02
-8.30989063e-01 4.40852970e-01 -3.85072678e-01 8.04377571e-02
1.20335329e+00 -3.94874036e-01 -1.31145269e-02 -6.87346339e-01
-7.10740209e-01 1.16292901e-01 6.69232368e-01 5.63534558e-01
7.54468560e-01 -1.73719358e+00 -7.05145478e-01 7.11857080e-02
1.89739063e-01 -3.61558646e-01 3.08371305e-01 7.78157473e-01
-9.04963672e-01 -5.05186990e-02 -4.65119272e-01 -7.17299938e-01
-1.36834872e+00 5.91204286e-01 -9.87109244e-02 1.06943130e-01
-9.49345529e-01 5.28816760e-01 8.92605409e-02 5.09224474e-01
1.42585650e-01 -2.78348982e-01 -3.57983351e-01 2.12219909e-01
8.54244828e-01 4.80033159e-01 -3.17844123e-01 -7.35921502e-01
-2.56701767e-01 5.34949064e-01 3.19419533e-01 -4.65802878e-01
1.42383814e+00 -3.85194510e-01 -1.13286972e-02 6.23571396e-01
1.39211583e+00 1.18027620e-01 -1.51552069e+00 -6.88645467e-02
6.60820603e-02 -1.03397727e+00 -3.02678883e-01 -8.29999298e-02
-9.92431104e-01 1.92192063e-01 2.16014788e-01 5.46178043e-01
1.35658717e+00 -7.53101632e-02 9.91500437e-01 -4.90737781e-02
4.41369087e-01 -1.08376026e+00 5.17564833e-01 2.01677203e-01
1.06281543e+00 -6.88397288e-01 1.93910748e-02 -4.39185202e-01
-6.84875548e-01 1.18535221e+00 1.55564874e-01 -4.40577954e-01
6.35101080e-01 2.87973225e-01 -4.02192414e-01 -2.27221578e-01
-7.35889494e-01 1.66629031e-02 1.94312990e-01 1.87816694e-01
2.03956291e-01 7.92638510e-02 -4.61221218e-01 4.24946547e-01
-3.16905051e-01 3.50360632e-01 7.43722022e-01 1.02442408e+00
-5.96639991e-01 -5.11061251e-01 -4.12335575e-01 1.74018919e-01
-1.49498180e-01 3.37237298e-01 -3.72395098e-01 2.67411888e-01
-1.52901798e-01 9.76582527e-01 1.46383598e-01 -7.82630742e-01
4.39869672e-01 -3.53750080e-01 6.42589271e-01 -3.03263724e-01
-2.95741230e-01 2.60449439e-01 4.70785312e-02 -9.21978772e-01
-8.37112248e-01 -1.01777983e+00 -1.01411688e+00 -5.80212712e-01
2.01201677e-01 1.19991325e-01 2.80621409e-01 8.61028433e-01
1.24340422e-01 4.48548675e-01 1.06139266e+00 -8.41670096e-01
2.18244880e-01 -4.64375615e-01 -7.24163890e-01 7.45302320e-01
3.47920984e-01 -5.33320904e-01 -1.61287606e-01 8.19553494e-01] | [10.572590827941895, 0.5877353549003601] |
b720c432-1d70-4f63-9480-8b89652e616a | learning-based-model-predictive-control-for | 1803.08287 | null | http://arxiv.org/abs/1803.08287v3 | http://arxiv.org/pdf/1803.08287v3.pdf | Learning-based Model Predictive Control for Safe Exploration | Learning-based methods have been successful in solving complex control tasks
without significant prior knowledge about the system. However, these methods
typically do not provide any safety guarantees, which prevents their use in
safety-critical, real-world applications. In this paper, we present a
learning-based model predictive control scheme that can provide provable
high-probability safety guarantees. To this end, we exploit regularity
assumptions on the dynamics in terms of a Gaussian process prior to construct
provably accurate confidence intervals on predicted trajectories. Unlike
previous approaches, we do not assume that model uncertainties are independent.
Based on these predictions, we guarantee that trajectories satisfy safety
constraints. Moreover, we use a terminal set constraint to recursively
guarantee the existence of safe control actions at every iteration. In our
experiments, we show that the resulting algorithm can be used to safely and
efficiently explore and learn about dynamic systems. | ['Andreas Krause', 'Torsten Koller', 'Matteo Turchetta', 'Felix Berkenkamp'] | 2018-03-22 | null | null | null | null | ['safe-exploration'] | ['robots'] | [ 1.59368053e-01 3.04628342e-01 -3.70559722e-01 1.77004606e-01
-8.00562918e-01 -6.94705069e-01 4.84627217e-01 5.88100553e-01
-1.31413132e-01 9.81782079e-01 -4.54455227e-01 -6.61007702e-01
-4.48783010e-01 -7.79555559e-01 -1.03248382e+00 -7.94063628e-01
-3.64484370e-01 4.37976539e-01 4.57239538e-01 3.53652537e-02
1.52306631e-01 3.23264956e-01 -1.12483025e+00 -4.20772582e-01
9.74380672e-01 9.63717937e-01 -1.89692438e-01 5.64341605e-01
6.20145440e-01 6.61047757e-01 -8.48720893e-02 2.25281239e-01
3.12628299e-01 -1.65555447e-01 -4.63835090e-01 1.00411601e-01
-2.82973796e-01 -3.68832558e-01 -2.44427994e-01 1.33041620e+00
1.23780720e-01 5.03654063e-01 5.01603067e-01 -1.65035903e+00
2.32446715e-01 3.36030006e-01 -3.20693493e-01 -3.45808089e-01
1.02556802e-01 4.11195993e-01 6.43801093e-01 -2.73000091e-01
1.67811662e-01 9.07047033e-01 6.17808700e-01 5.64534664e-01
-1.36182058e+00 -5.82188487e-01 4.75052327e-01 -8.36486295e-02
-1.34154928e+00 -2.34728366e-01 4.45402205e-01 -5.29828548e-01
5.27266502e-01 2.58190423e-01 4.69525307e-01 8.26223612e-01
6.15587056e-01 5.85178137e-01 8.92553389e-01 -2.42831916e-01
6.88730836e-01 5.99516183e-02 -7.65572041e-02 4.83142853e-01
4.95955378e-01 5.62539935e-01 1.56743824e-02 -5.16596317e-01
6.01210833e-01 2.28204310e-01 -2.87486404e-01 -6.42128885e-01
-9.77349460e-01 9.01364148e-01 -5.65476567e-02 -1.67581797e-01
-2.75489748e-01 4.60677236e-01 3.35912287e-01 1.84355289e-01
3.16560626e-01 2.89612412e-01 -3.74655008e-01 -1.86092481e-02
-4.93456036e-01 5.36033809e-01 8.18150640e-01 1.03207278e+00
3.43269736e-01 3.06724101e-01 -2.43668661e-01 -1.22597720e-02
3.13125789e-01 7.34037399e-01 -2.79735595e-01 -1.16700172e+00
2.47243971e-01 -3.39992084e-02 8.84606540e-01 -7.69980609e-01
-1.03890277e-01 -1.78155258e-01 -6.00907326e-01 5.75243115e-01
3.30059588e-01 -4.53208745e-01 -8.71616602e-01 1.73917603e+00
3.58388454e-01 6.58617675e-01 6.86290935e-02 6.07599258e-01
-7.65860558e-01 8.90646338e-01 -1.04337953e-01 -6.31536663e-01
6.49681330e-01 -5.17957985e-01 -6.91680312e-01 2.99603343e-02
5.06191015e-01 -1.91015601e-01 8.08946967e-01 6.38028264e-01
-1.05906749e+00 -8.33381042e-02 -9.64874625e-01 8.02674055e-01
2.32992560e-01 -3.11611086e-01 1.79838911e-01 5.45632124e-01
-6.34425581e-01 6.70564175e-01 -1.36943281e+00 7.17871115e-02
2.11849362e-01 3.13005954e-01 -4.67541516e-02 1.44916534e-01
-8.76552641e-01 9.08889532e-01 6.84687018e-01 1.14983700e-01
-1.65761685e+00 -8.07917774e-01 -8.69613886e-01 1.75757408e-02
1.14657080e+00 -4.01561350e-01 1.60311294e+00 -2.05677450e-01
-1.59778845e+00 -1.20310836e-01 -1.15549721e-01 -7.99375355e-01
7.36263216e-01 -4.35399711e-01 -6.04280047e-02 4.94067483e-02
-3.11015338e-01 -1.12283304e-01 9.98366058e-01 -1.42956662e+00
-9.28411841e-01 -2.41535399e-02 2.63034523e-01 -1.28283709e-01
-1.56142667e-01 -3.36417764e-01 -6.82551414e-02 -2.11626038e-01
-3.31576496e-01 -1.27890420e+00 -8.49234641e-01 1.69433787e-01
-4.15577471e-01 -7.59136379e-02 8.41256082e-01 -4.25770551e-01
1.11348283e+00 -1.94016731e+00 -5.33096259e-03 6.90511167e-01
-6.87056184e-02 4.08768624e-01 3.39941233e-01 6.72864318e-01
3.22260290e-01 1.23602442e-01 -5.22298932e-01 -1.89758569e-01
1.21918052e-01 4.66537446e-01 -8.93974900e-01 7.83707380e-01
1.82633445e-01 3.35319072e-01 -8.89180839e-01 -1.32560775e-01
4.45236117e-01 1.79696053e-01 -5.51755786e-01 2.32235715e-01
-7.22778380e-01 7.80049980e-01 -9.07230973e-01 -9.07274634e-02
4.56626892e-01 1.28136292e-01 1.79621920e-01 5.71934938e-01
-9.27293822e-02 -1.43796325e-01 -1.12087166e+00 1.01112831e+00
-7.32276440e-01 1.15709521e-01 3.38830113e-01 -1.07310510e+00
5.67200541e-01 4.33434069e-01 4.38473195e-01 -9.43356454e-02
2.50465155e-01 3.51117440e-02 -2.65009880e-01 -8.77416208e-02
-6.87601939e-02 -3.49823624e-01 -1.59025759e-01 1.91432685e-01
-4.10046995e-01 -4.50725853e-01 -9.75915641e-02 -1.73285231e-01
9.66762960e-01 -7.25098625e-02 2.04341367e-01 -3.56340855e-01
6.01800442e-01 7.04969615e-02 9.75620687e-01 6.20786846e-01
3.04586254e-03 4.70733084e-03 7.70994842e-01 7.81383812e-02
-1.06763363e+00 -1.13499165e+00 1.02121405e-01 4.52643424e-01
1.70271471e-01 -2.26953596e-01 -5.75147808e-01 -4.75940734e-01
5.26207834e-02 1.08917570e+00 -5.34063995e-01 -4.18055743e-01
-4.02001470e-01 -2.47569159e-01 8.18462297e-02 6.19799733e-01
-9.83104259e-02 -2.90432632e-01 -8.63767147e-01 4.06629533e-01
4.06657428e-01 -9.17960882e-01 -5.00424206e-01 1.62588596e-01
-7.11235225e-01 -1.29713261e+00 -2.64144063e-01 -3.99525702e-01
8.92905772e-01 -2.26957634e-01 4.84890461e-01 -2.03412727e-01
1.58793014e-02 5.86311281e-01 2.08446041e-01 -5.71779013e-01
-5.81613421e-01 -3.59470010e-01 2.05666691e-01 -1.42380118e-01
-5.57489395e-01 -3.35099161e-01 -1.57921731e-01 2.21606851e-01
-8.38898838e-01 -1.14353232e-01 9.92274880e-02 8.00035417e-01
9.11863983e-01 8.77728343e-01 5.29217839e-01 -7.40625203e-01
7.10733354e-01 -3.18650067e-01 -1.62214744e+00 3.52201074e-01
-7.98801184e-01 2.14905664e-01 1.03300691e+00 -5.30551016e-01
-8.41461897e-01 2.97721952e-01 2.10480154e-01 -8.26421618e-01
-1.40510842e-01 5.17733514e-01 -1.40320882e-01 3.42387334e-02
1.25999928e-01 1.52185962e-01 2.73757219e-01 -1.60806611e-01
1.51039302e-01 1.29745126e-01 5.61968386e-01 -9.39500630e-01
1.19327033e+00 3.25113058e-01 5.06514907e-01 -4.89492178e-01
-7.12785482e-01 -2.20792353e-01 -2.94158906e-01 -2.31145620e-01
6.43538356e-01 -6.96367919e-01 -1.28094566e+00 1.76828474e-01
-6.64885581e-01 -6.90180659e-01 -1.21040881e-01 4.86758262e-01
-1.17144144e+00 1.80628195e-01 -4.05758291e-01 -1.77743602e+00
9.96942297e-02 -9.27003026e-01 6.38874471e-01 1.57005399e-01
-6.83314130e-02 -1.20729518e+00 1.74703404e-01 -2.53118217e-01
2.91143715e-01 7.57391453e-01 7.85108745e-01 -5.73540628e-01
-5.19907355e-01 -3.49653661e-01 3.92476410e-01 2.33593628e-01
4.98669408e-02 5.08762896e-02 -4.66364980e-01 -5.33407390e-01
1.93215206e-01 -1.82130009e-01 5.52273691e-01 2.43559033e-01
1.34415829e+00 -8.68725419e-01 -5.67156971e-01 -4.60636541e-02
1.37640560e+00 4.85437691e-01 2.16659978e-01 -6.68370128e-02
3.63420635e-01 6.55536056e-01 1.11393619e+00 8.51382613e-01
7.13963434e-02 3.93562675e-01 6.74043953e-01 4.07948852e-01
8.91550183e-01 -4.16706622e-01 4.61250693e-01 9.96088013e-02
1.56954125e-01 -2.03309819e-01 -1.14496779e+00 7.84991443e-01
-2.19559932e+00 -8.37665141e-01 4.72589210e-02 2.91109300e+00
1.02517641e+00 4.56574202e-01 1.06848687e-01 1.21823169e-01
7.40756512e-01 -3.62830907e-01 -8.22996616e-01 -2.67016053e-01
6.44447029e-01 -3.11221611e-02 8.95189404e-01 8.16738605e-01
-1.18633723e+00 4.76301968e-01 6.28391933e+00 8.12047124e-01
-9.09136832e-01 -1.37430087e-01 5.40608227e-01 4.85486761e-02
-3.10144246e-01 1.81099147e-01 -6.71083093e-01 4.75684583e-01
1.20345354e+00 -7.40716934e-01 5.11733592e-01 1.06122267e+00
5.47901273e-01 -1.22696668e-01 -1.14748931e+00 3.99816096e-01
-6.19135618e-01 -8.64715159e-01 -5.23003042e-01 5.05462997e-02
8.10980678e-01 -5.23559570e-01 3.56919944e-01 2.10825562e-01
7.71072388e-01 -1.12352645e+00 7.31596529e-01 5.68851054e-01
4.90717232e-01 -1.48284984e+00 5.65393031e-01 7.73253322e-01
-1.00137818e+00 -4.31546003e-01 -6.30765781e-02 -4.62459847e-02
4.91436392e-01 6.39734626e-01 -6.92082047e-01 5.55559158e-01
8.43479186e-02 4.39181447e-01 1.37139738e-01 1.26494312e+00
-4.57300067e-01 8.74414921e-01 -6.30170822e-01 1.31125614e-01
2.66928375e-01 -5.46407215e-02 6.53965175e-01 6.71036899e-01
2.55948395e-01 -5.86926192e-02 8.58160615e-01 7.82003045e-01
5.05468249e-01 -5.16504943e-01 -7.42706060e-01 -1.17583372e-01
4.99603301e-01 6.95660412e-01 -5.65433800e-01 -2.28329226e-01
-2.81993747e-01 4.58095253e-01 -8.59538913e-02 3.81591767e-01
-1.19048560e+00 -2.47669920e-01 8.15588176e-01 -2.81682778e-02
3.91260773e-01 -7.31066287e-01 -3.17659259e-01 -7.23168552e-01
7.52453553e-03 -6.71313345e-01 2.43176863e-01 -1.56596377e-01
-1.20001256e+00 7.37634674e-03 2.65143722e-01 -1.04416215e+00
-6.79578125e-01 -3.64137143e-01 -8.55268300e-01 7.79762805e-01
-1.19664514e+00 -7.59604990e-01 4.25768763e-01 5.99318683e-01
3.19429159e-01 3.57303947e-01 5.57463169e-01 -3.29468489e-01
-6.99716568e-01 3.11089102e-02 3.11617523e-01 -2.15969518e-01
5.60889721e-01 -1.29864800e+00 7.14271888e-02 1.07481182e+00
-4.62020159e-01 4.77454513e-01 1.25241172e+00 -8.45955312e-01
-1.60568607e+00 -1.60859895e+00 1.19169913e-01 -3.47637624e-01
1.05502999e+00 -1.48977771e-01 -9.73590553e-01 8.10472310e-01
-1.58980951e-01 2.09459156e-01 5.71490005e-02 -8.04327205e-02
-3.19221281e-02 4.26450446e-02 -1.13175082e+00 5.85954785e-01
6.29780412e-01 -2.39203811e-01 -2.70822197e-01 2.77916342e-01
8.81040096e-01 -3.99198979e-01 -9.39232230e-01 7.43728638e-01
7.55465701e-02 -1.77676409e-01 7.12370098e-01 -8.55117440e-01
-2.24274293e-01 -4.56525415e-01 5.59751596e-03 -1.31213117e+00
-3.65169942e-02 -9.81716812e-01 -6.10372245e-01 9.32342291e-01
4.37821925e-01 -9.37413633e-01 5.30594707e-01 1.04849207e+00
-2.76989520e-01 -7.48698950e-01 -1.15677249e+00 -1.40270245e+00
3.74824613e-01 -6.27237082e-01 3.04271460e-01 4.56214786e-01
2.72497803e-01 -3.05970609e-01 -6.88225865e-01 7.93357909e-01
8.44954848e-01 1.17431052e-01 5.16435385e-01 -8.81488323e-01
-4.56611454e-01 -3.23834300e-01 -3.03487126e-02 -4.50055689e-01
5.49763143e-01 -1.07709467e-01 7.46067941e-01 -1.20292068e+00
6.38452023e-02 -6.37075305e-01 -3.00350547e-01 5.33420682e-01
-2.79940933e-01 -3.80893916e-01 9.55691636e-02 -1.79399908e-01
-7.25847721e-01 8.75007212e-01 8.46350372e-01 5.02312817e-02
-2.25734606e-01 5.35951376e-01 -2.42971793e-01 7.33886898e-01
1.13225710e+00 -3.44103038e-01 -9.41169322e-01 1.23219870e-01
1.66258160e-02 4.51930434e-01 3.65009516e-01 -1.00789487e+00
1.61491662e-01 -9.66794610e-01 -3.27368200e-01 -5.90787053e-01
1.65306002e-01 -1.18332291e+00 2.69525409e-01 1.04848564e+00
-4.35426205e-01 -3.27778965e-01 3.81928116e-01 1.22055423e+00
3.12458724e-02 -1.49868995e-01 9.29357409e-01 3.94729882e-01
-3.97322118e-01 6.22174799e-01 -7.84952819e-01 -4.10836153e-02
1.61213899e+00 5.45537829e-01 1.29846171e-01 -6.60149992e-01
-7.79263020e-01 8.63045096e-01 4.81715590e-01 1.44430950e-01
4.39023286e-01 -1.00653923e+00 -3.51429909e-01 -5.10634705e-02
-1.75544158e-01 1.21416837e-01 1.97264791e-01 8.69934440e-01
5.15069962e-02 6.09466314e-01 1.84573382e-01 -5.09233654e-01
-9.46055949e-01 1.26628399e+00 2.46794492e-01 -1.53130755e-01
-5.27407587e-01 1.43626273e-01 1.98571801e-01 -8.60930309e-02
2.94591874e-01 -4.58177507e-01 3.44306409e-01 -6.45934105e-01
6.44400775e-01 4.51321363e-01 -2.09931642e-01 1.76760275e-02
-2.16589659e-01 1.97026461e-01 1.69641688e-01 -4.34144825e-01
1.16450799e+00 -1.68460846e-01 2.56987005e-01 5.87021470e-01
6.28391862e-01 -2.20936574e-02 -1.89368832e+00 3.93977128e-02
2.37013549e-01 -4.25955951e-01 -6.41788170e-02 -4.84863669e-01
-8.32516193e-01 7.19834387e-01 2.36536175e-01 2.35957414e-01
1.00185466e+00 -3.24339420e-01 7.04372168e-01 5.02061188e-01
7.78935552e-01 -1.00514400e+00 -3.62395018e-01 6.13247752e-01
6.52157307e-01 -8.38422239e-01 -3.25452507e-01 -5.13996422e-01
-5.62930226e-01 8.86649311e-01 6.04455113e-01 -4.51104224e-01
6.98037088e-01 6.31522059e-01 -5.95276296e-01 5.97901881e-01
-1.20312214e+00 -7.86554664e-02 4.30238061e-02 7.19088912e-01
-1.93487406e-01 1.40420571e-02 -1.30804017e-01 4.29396957e-01
3.00514489e-01 3.98977324e-02 5.76971173e-01 1.18642807e+00
-7.72392035e-01 -1.03602910e+00 -6.21514261e-01 1.42453954e-01
-4.09144163e-01 4.53917533e-01 1.27434328e-01 6.98946476e-01
-4.05844778e-01 1.17230678e+00 -2.20063791e-01 -1.46477059e-01
2.11227641e-01 3.68445227e-03 4.22545373e-01 -5.90702832e-01
9.66587290e-02 1.33613199e-01 1.97263822e-01 -6.77831709e-01
2.16215685e-01 -6.29530311e-01 -1.38727653e+00 -5.05130112e-01
-3.28914672e-01 3.50078553e-01 3.11461210e-01 1.01582849e+00
2.80596912e-01 4.59304333e-01 7.73219049e-01 -4.34628159e-01
-1.23174345e+00 -4.15117294e-01 -4.57829267e-01 -2.58085549e-01
5.03762007e-01 -8.43450069e-01 -3.21933210e-01 -2.26534214e-02] | [4.821493148803711, 2.2602005004882812] |
c373f7d5-8c52-4633-8c1f-7441a24046a5 | pretraining-the-noisy-channel-model-for-task | 2103.10518 | null | https://arxiv.org/abs/2103.10518v1 | https://arxiv.org/pdf/2103.10518v1.pdf | Pretraining the Noisy Channel Model for Task-Oriented Dialogue | Direct decoding for task-oriented dialogue is known to suffer from the explaining-away effect, manifested in models that prefer short and generic responses. Here we argue for the use of Bayes' theorem to factorize the dialogue task into two models, the distribution of the context given the response, and the prior for the response itself. This approach, an instantiation of the noisy channel model, both mitigates the explaining-away effect and allows the principled incorporation of large pretrained models for the response prior. We present extensive experiments showing that a noisy channel model decodes better responses compared to direct decoding and that a two stage pretraining strategy, employing both open-domain and task-oriented dialogue data, improves over randomly initialized models. | ['Phil Blunsom', 'Laura Rimell', 'Lei Yu', 'Qi Liu'] | 2021-03-18 | null | null | null | null | ['end-to-end-dialogue-modelling'] | ['natural-language-processing'] | [ 5.61890662e-01 7.40145445e-01 1.28961533e-01 -8.96346092e-01
-1.17489803e+00 -6.46313787e-01 8.75387788e-01 -8.94662924e-03
-5.97311258e-01 9.67288673e-01 9.53862846e-01 -5.41293085e-01
1.44612908e-01 -4.28348631e-01 -2.27057844e-01 -2.79491931e-01
2.78345019e-01 7.99800098e-01 1.11191563e-01 -5.84683776e-01
3.94254416e-01 -2.06711128e-01 -8.21236789e-01 7.38207579e-01
4.61653739e-01 6.50830507e-01 3.44440341e-01 1.33323026e+00
-3.56873870e-01 1.01581144e+00 -1.05020082e+00 -4.86509532e-01
-1.27736777e-01 -7.99846470e-01 -1.32792842e+00 1.18811473e-01
-2.17740968e-01 -5.47826231e-01 -3.74493092e-01 6.37174904e-01
6.56565428e-01 1.94388449e-01 5.84246576e-01 -6.97952271e-01
-5.73765099e-01 7.69326031e-01 9.97913629e-02 8.06249306e-02
9.47961509e-01 9.09696072e-02 1.16020370e+00 -3.23428631e-01
5.87717593e-01 1.51531625e+00 7.14129984e-01 1.10547602e+00
-1.45448852e+00 -1.97243214e-01 -1.63701084e-02 -3.31656545e-01
-7.86421716e-01 -6.86996818e-01 4.40164179e-01 -3.54631186e-01
1.25438881e+00 3.25254053e-01 3.14266086e-01 1.63984609e+00
1.67646676e-01 8.86615753e-01 1.59693050e+00 -5.81243634e-01
1.59468755e-01 3.76785189e-01 2.71736234e-01 1.59280270e-01
-4.25092459e-01 1.77152887e-01 -6.43246531e-01 -5.05640447e-01
5.50919354e-01 -5.66160560e-01 -4.60758179e-01 1.54763624e-01
-9.00412738e-01 1.09659779e+00 -1.37145743e-01 9.20409244e-03
-2.41205230e-01 9.04977918e-02 5.97824097e-01 7.75155306e-01
6.85812473e-01 5.97076952e-01 -8.40401888e-01 -7.65569389e-01
-5.83711028e-01 3.16930085e-01 1.59973419e+00 9.91750062e-01
5.56424379e-01 -2.12781668e-01 -3.69513959e-01 1.01167929e+00
5.79572141e-01 2.46055648e-01 3.89035642e-01 -1.04398811e+00
5.94901264e-01 8.05274323e-02 2.56973416e-01 -5.56246042e-01
-4.92505640e-01 -1.17927969e-01 -3.03483754e-01 -4.00717497e-01
8.71923923e-01 -7.14941323e-01 -5.50282598e-01 2.02882719e+00
3.76958959e-02 -5.94904900e-01 5.14535010e-01 7.49320745e-01
8.02643001e-01 7.09120691e-01 1.80815890e-01 -3.70762110e-01
1.23032451e+00 -8.23854566e-01 -9.68342662e-01 -5.58734119e-01
7.92714238e-01 -9.41241801e-01 1.04741871e+00 3.98814917e-01
-1.14875889e+00 -2.90442705e-01 -7.89676905e-01 -2.15018123e-01
6.73535392e-02 -2.63618618e-01 7.75179803e-01 8.11548531e-01
-1.15454197e+00 3.66227508e-01 -2.50111133e-01 -4.66949612e-01
-1.39287189e-01 4.08506811e-01 -2.22579122e-01 -1.62745621e-02
-1.43387997e+00 1.18509233e+00 4.07106131e-02 9.71777290e-02
-9.45300579e-01 5.58316484e-02 -7.99736023e-01 4.74648103e-02
2.29285628e-01 -4.81345206e-01 1.97177804e+00 -1.12940991e+00
-2.11415696e+00 7.82521427e-01 -2.84348249e-01 -3.18122059e-01
4.26305294e-01 -2.45395869e-01 5.05266599e-02 1.01731857e-02
-2.29868576e-01 5.39843976e-01 7.85833716e-01 -1.12704504e+00
-3.27426583e-01 -2.45930180e-01 4.16267186e-01 4.53504086e-01
-8.95790383e-02 1.98215261e-01 -2.34302655e-01 -2.75337368e-01
-5.93731813e-02 -9.25991237e-01 -2.53718734e-01 -6.73772812e-01
-3.60154599e-01 -3.53590637e-01 3.88206571e-01 -8.65489006e-01
1.09588706e+00 -1.87640011e+00 1.31712437e-01 8.08412954e-02
1.72628030e-01 -1.70072302e-01 -2.19914883e-01 9.33169842e-01
9.90911052e-02 1.14031248e-02 -2.38005985e-02 -4.23594326e-01
2.37107083e-01 6.14369512e-01 -1.59524873e-01 2.89364278e-01
2.65944656e-02 7.27642000e-01 -8.70398879e-01 -2.86506545e-02
-2.76523143e-01 3.35235655e-01 -6.28444254e-01 6.64708018e-01
-4.17533129e-01 6.73099995e-01 -4.16255176e-01 -2.48822682e-02
2.91058034e-01 -1.89494297e-01 7.73194909e-01 4.89533871e-01
9.42724198e-02 1.23707116e+00 -7.35344231e-01 1.79433084e+00
-5.25805414e-01 6.70149744e-01 6.87475502e-01 -6.81412995e-01
8.78426194e-01 8.41402829e-01 -8.78223777e-02 -5.24576604e-01
3.00545841e-01 3.34699035e-01 3.28681499e-01 -6.31403983e-01
6.55523658e-01 -6.05883479e-01 -4.56972659e-01 9.42709982e-01
2.91780412e-01 -2.96339869e-01 -3.48842114e-01 5.76033056e-01
1.06550562e+00 1.86998501e-01 1.95160270e-01 -2.56356150e-01
3.39830101e-01 -7.56227411e-03 2.11477727e-01 1.17746985e+00
-2.07989916e-01 4.73356813e-01 8.31931412e-01 -1.47362888e-01
-9.74055052e-01 -3.58924121e-01 -4.21568491e-02 1.55161655e+00
-1.75910190e-01 -5.79635441e-01 -9.63926673e-01 -8.44314635e-01
-1.81733236e-01 8.65568697e-01 -4.62497830e-01 -6.26886338e-02
-3.91032755e-01 -6.92880929e-01 8.22028637e-01 3.73795182e-01
4.65667136e-02 -7.94803858e-01 -5.72985649e-01 3.90680522e-01
-7.01905012e-01 -9.42547739e-01 -4.13806766e-01 7.87086546e-01
-7.33412564e-01 -6.42073870e-01 -6.30871832e-01 -4.69618618e-01
3.15661192e-01 1.22061251e-02 1.17714453e+00 2.21238196e-01
3.93644720e-01 7.14240789e-01 -5.62128186e-01 -2.53913850e-01
-1.12682164e+00 6.97668735e-03 -3.33209962e-01 -2.96478420e-01
5.92710614e-01 -1.07936718e-01 -2.76745677e-01 2.21581072e-01
-5.28149128e-01 6.83616698e-02 4.21014726e-01 1.34109104e+00
-3.71197432e-01 -5.17385483e-01 7.17670560e-01 -1.31926227e+00
1.37046695e+00 -5.30557454e-01 5.63982502e-02 6.43657073e-02
-5.93836844e-01 2.63172328e-01 4.12505627e-01 -2.67617673e-01
-1.40858412e+00 -1.52209938e-01 -6.16147399e-01 5.68903089e-01
-4.53536898e-01 4.64473903e-01 -1.10118553e-01 3.08981650e-02
7.66555190e-01 9.68047976e-02 2.58399427e-01 -5.86281955e-01
4.02326941e-01 1.21452069e+00 2.39575237e-01 -9.97556329e-01
1.69401824e-01 -6.25711307e-02 -6.97964907e-01 -8.36664021e-01
-6.95245087e-01 -5.15461624e-01 -3.44525874e-01 -2.23651960e-01
8.31099987e-01 -7.90562749e-01 -4.32341725e-01 2.53042191e-01
-1.61244178e+00 -7.83772290e-01 -1.65138751e-01 5.18294334e-01
-7.61371374e-01 5.19330800e-01 -1.01387823e+00 -1.09505057e+00
3.55256945e-02 -1.12870157e+00 9.23864663e-01 -2.79146861e-02
-6.91188455e-01 -1.27872694e+00 2.98561957e-02 5.54833233e-01
6.40997350e-01 -4.19542700e-01 9.84265804e-01 -1.40170252e+00
-1.99202791e-01 -2.53200203e-01 -1.21693432e-01 3.90756577e-01
-3.42228293e-01 -5.94103992e-01 -1.35009360e+00 -1.20422855e-01
5.93841553e-01 -9.34383571e-01 6.44779205e-01 8.59607682e-02
3.33366513e-01 -4.99341697e-01 3.84694338e-02 3.61115560e-02
1.02974606e+00 8.01967084e-02 6.15425050e-01 8.03411081e-02
2.71272719e-01 9.82720017e-01 3.38040948e-01 5.96940875e-01
7.54875779e-01 5.57814479e-01 7.66022876e-02 2.14627355e-01
1.52309537e-01 -3.29701573e-01 4.89972800e-01 9.39650536e-01
2.12624297e-01 -6.06985569e-01 -5.68769991e-01 4.70306814e-01
-1.76544297e+00 -6.76564634e-01 -3.93142641e-01 2.07860446e+00
1.35027552e+00 1.90981120e-01 1.21462531e-01 -1.03180423e-01
4.06289458e-01 2.67211825e-01 1.60361510e-02 -1.15563464e+00
6.81430027e-02 1.78375795e-01 4.19690579e-01 1.05187678e+00
-6.61162138e-01 8.78184080e-01 7.98963165e+00 5.21608114e-01
-6.65636897e-01 5.03638923e-01 5.03578842e-01 3.94286335e-01
-5.43574333e-01 3.13516676e-01 -7.43320584e-01 2.65345663e-01
1.45468962e+00 1.35215044e-01 5.76629817e-01 4.64500904e-01
2.45438769e-01 -3.63804162e-01 -1.39247000e+00 4.06093836e-01
4.99205552e-02 -7.93538094e-01 -2.20478237e-01 7.59365335e-02
2.54286170e-01 -1.48934454e-01 -2.66456455e-01 4.72534239e-01
6.79194152e-01 -9.87757564e-01 6.82245374e-01 1.14575572e-01
5.42725861e-01 -2.50917137e-01 8.50617886e-01 7.63491094e-01
-4.27612126e-01 -9.50485319e-02 -3.38800222e-01 -6.10717177e-01
1.44978419e-01 -8.57914053e-03 -1.09989667e+00 1.21849008e-01
1.10083759e-01 1.36210620e-01 -3.74972336e-02 4.50553954e-01
-4.90381569e-01 7.47067690e-01 -1.89520434e-01 -2.65935808e-01
4.22731161e-01 2.46355813e-02 5.55423617e-01 1.57069874e+00
-2.47340098e-01 3.56841683e-01 3.21354598e-01 6.18683219e-01
1.26443401e-01 2.12636422e-02 -5.83098769e-01 -7.30065554e-02
3.67910981e-01 9.37768459e-01 -3.67321312e-01 -3.53474766e-01
-5.37289441e-01 1.12132108e+00 1.38260722e-01 4.06973094e-01
-4.42507625e-01 -1.98295176e-01 2.27011979e-01 -2.37187311e-01
1.94926292e-01 -2.44417340e-01 -4.30716336e-01 -1.03395438e+00
-2.93157309e-01 -1.26186800e+00 2.67290324e-01 -6.08960092e-01
-1.09119213e+00 5.92655838e-01 -8.36574435e-02 -4.57980573e-01
-7.04014421e-01 -6.14539146e-01 -3.99752051e-01 1.36212647e+00
-1.49698973e+00 -7.71440327e-01 4.29124758e-02 5.20066738e-01
8.49940836e-01 1.52422324e-01 1.37754464e+00 1.13042016e-02
-1.86665878e-01 5.03288329e-01 -6.73761815e-02 8.50646719e-02
9.83335435e-01 -1.56223500e+00 2.06020966e-01 1.91072360e-01
-2.12179199e-01 9.57233369e-01 1.02159619e+00 -5.38354039e-01
-1.40258420e+00 -3.65512341e-01 1.38689160e+00 -8.13672960e-01
5.55178821e-01 -6.23320103e-01 -8.66742492e-01 7.45337844e-01
7.43182361e-01 -8.25378537e-01 1.10246146e+00 4.08491999e-01
-3.56497169e-01 4.57963288e-01 -9.88329232e-01 1.37892962e-01
4.95921016e-01 -9.35071588e-01 -9.93403614e-01 4.90098298e-01
6.24045372e-01 -6.30967557e-01 -8.85536134e-01 -2.59401232e-01
6.57456875e-01 -1.02448797e+00 3.29151064e-01 -6.79080963e-01
4.12501305e-01 4.61663991e-01 -3.48650455e-01 -1.48355639e+00
-1.22476459e-01 -1.29337108e+00 7.70158172e-02 9.71517444e-01
6.43732071e-01 -5.49573302e-01 5.25403559e-01 9.60731745e-01
-1.53546289e-01 -4.10284638e-01 -7.80168772e-01 -2.88646817e-01
2.34407857e-01 -5.96778095e-01 8.61913860e-02 6.93441868e-01
5.98996937e-01 1.08240044e+00 -6.96823955e-01 -2.13652790e-01
1.31751344e-01 -3.71063650e-01 8.09979916e-01 -1.06229579e+00
-7.54045665e-01 -1.03958420e-01 9.74044204e-02 -2.13418269e+00
1.55255377e-01 -4.58833337e-01 6.86656594e-01 -1.33254397e+00
1.07607469e-01 -3.86501193e-01 2.95960367e-01 1.78820118e-01
-1.57364830e-01 -6.70562759e-02 1.11424066e-01 1.95465043e-01
-6.37260258e-01 3.17497313e-01 1.19191372e+00 1.77616924e-01
-1.38069645e-01 1.04502529e-01 -1.09634876e+00 5.24060309e-01
5.11118650e-01 -6.37761891e-01 -4.89722282e-01 -6.15905046e-01
3.91343355e-01 6.12005711e-01 9.77704301e-02 -1.95585117e-01
3.55490237e-01 9.91643295e-02 2.20798887e-02 -6.71715140e-02
5.80443621e-01 -5.12438416e-01 -3.62976223e-01 1.60016820e-01
-1.13908601e+00 -1.98473275e-01 5.76391928e-02 9.09325659e-01
-5.03858812e-02 -4.98456955e-01 5.41486382e-01 -6.46928668e-01
-2.82223046e-01 -5.12696981e-01 -9.75786686e-01 2.95306772e-01
2.88960725e-01 -2.55449623e-01 -2.93405652e-01 -1.08756042e+00
-8.71995986e-01 2.37123042e-01 1.41422823e-01 2.27142900e-01
4.30463403e-01 -7.17416763e-01 -6.97424889e-01 1.83904067e-01
-1.64297163e-01 -2.39694044e-01 -1.07316412e-01 9.94456887e-01
-1.74172699e-01 6.17587447e-01 1.98980585e-01 -4.03637975e-01
-1.13672614e+00 1.91504788e-02 3.21874738e-01 -4.31945235e-01
-5.21279275e-01 1.24230754e+00 2.24721089e-01 -7.97936738e-01
4.74247307e-01 -1.19138859e-01 -1.79881707e-01 -8.86830539e-02
4.82556969e-01 -9.90714878e-03 -6.51850030e-02 -5.91683090e-01
6.81605339e-02 -2.56167293e-01 -2.92436987e-01 -6.19668961e-01
1.02324533e+00 -6.02933586e-01 -9.11753029e-02 7.02432871e-01
9.31426585e-01 2.36197338e-02 -1.26267302e+00 -3.83459598e-01
1.42257944e-01 -4.42846656e-01 -2.72399038e-02 -1.27171874e+00
-1.46975994e-01 1.03378737e+00 5.57331294e-02 7.31924772e-01
6.17458045e-01 -1.54891953e-01 8.66337836e-01 4.79318619e-01
3.55279356e-01 -1.24955547e+00 6.81999698e-02 9.83083487e-01
6.46514714e-01 -1.05916333e+00 -2.57647246e-01 -3.08026940e-01
-1.05233645e+00 1.21475303e+00 4.65513110e-01 1.72185913e-01
3.27667296e-01 3.88819605e-01 5.14550984e-01 -2.48408079e-01
-1.24798214e+00 1.53797632e-02 -2.56778121e-01 7.84416020e-01
8.73365521e-01 2.40657143e-02 -4.76120919e-01 8.43780518e-01
-1.94180205e-01 -2.46630609e-01 8.12045872e-01 9.77426112e-01
-5.10940850e-01 -1.33931625e+00 -3.76489252e-01 1.56583473e-01
-7.54133165e-01 -2.36876115e-01 -1.14868557e+00 5.57450533e-01
-3.07300329e-01 1.50499821e+00 -2.53357023e-01 -4.17845577e-01
1.34717330e-01 6.17762685e-01 4.52295095e-01 -9.51953292e-01
-1.09010947e+00 5.28889000e-01 9.79590178e-01 -3.50058615e-01
-3.48225147e-01 -3.94705534e-01 -9.50785398e-01 -3.10757339e-01
-8.09445083e-01 6.30540907e-01 6.45523667e-01 1.26426315e+00
1.16509363e-01 4.41499978e-01 6.08121574e-01 -6.59537256e-01
-1.38147199e+00 -1.38430393e+00 -4.87224609e-01 3.24023753e-01
3.64153236e-01 -2.30402529e-01 -3.87244612e-01 -1.19518720e-01] | [12.876802444458008, 7.950674533843994] |
5eadf271-65dd-4f9f-99e1-8a84858e4cac | inertial-navigation-meets-deep-learning-a | 2307.00014 | null | https://arxiv.org/abs/2307.00014v1 | https://arxiv.org/pdf/2307.00014v1.pdf | Inertial Navigation Meets Deep Learning: A Survey of Current Trends and Future Directions | Inertial sensing is used in many applications and platforms, ranging from day-to-day devices such as smartphones to very complex ones such as autonomous vehicles. In recent years, the development of machine learning and deep learning techniques has increased significantly in the field of inertial sensing. This is due to the development of efficient computing hardware and the accessibility of publicly available sensor data. These data-driven approaches are used to empower model-based navigation and sensor fusion algorithms. This paper provides an in-depth review of those deep learning methods. We examine separately, each vehicle operation domain including land, air, and sea. Each domain is divided into pure inertial advances and improvements based on filter parameters learning. In addition, we review deep learning approaches for calibrating and denoising inertial sensors. Throughout the paper, we discuss these trends and future directions. We also provide statistics on the commonly used approaches to illustrate their efficiency and stimulate further research in deep learning embedded in inertial navigation and fusion. | ['Itzik Klein', 'Nadav Cohen'] | 2023-06-22 | null | null | null | null | ['autonomous-vehicles'] | ['computer-vision'] | [-1.44861177e-01 -2.59095818e-01 -2.48691723e-01 -4.53145832e-01
-6.46628320e-01 -1.75233513e-01 4.29895252e-01 -2.35054210e-01
-5.79551637e-01 9.13561344e-01 3.14887226e-01 -4.23349589e-01
2.04991564e-01 -9.30398941e-01 -8.79446924e-01 -6.94598734e-01
-6.93292022e-02 -1.37283653e-02 -2.15912253e-01 -4.18521523e-01
-1.20666355e-01 4.57828194e-01 -1.83611405e+00 -4.59518462e-01
7.88077652e-01 1.21463823e+00 8.00164789e-02 8.40817034e-01
1.86430290e-01 4.24549460e-01 -3.96097481e-01 -9.99567658e-02
3.98130447e-01 -1.58879861e-01 2.62571406e-02 -3.03156108e-01
4.13331062e-01 -4.06841874e-01 -4.23419744e-01 1.02528834e+00
9.12421763e-01 1.73511282e-01 3.23504955e-01 -9.12552595e-01
-2.26226687e-01 2.72254407e-01 -3.43765885e-01 2.44760931e-01
3.38627607e-01 -3.80834311e-01 5.06905138e-01 -1.01234865e+00
1.14973016e-01 9.27617013e-01 1.30300331e+00 3.95772249e-01
-4.87902999e-01 -5.89624226e-01 -1.57657191e-01 1.25657246e-01
-1.13668561e+00 -6.40724480e-01 6.33975446e-01 -5.19035101e-01
1.22915602e+00 3.17824678e-03 8.21692526e-01 1.01932311e+00
7.81953573e-01 7.42544651e-01 7.25175619e-01 -3.56149107e-01
3.55433047e-01 -3.05140018e-01 1.40688587e-02 6.60588384e-01
6.82456732e-01 5.31257689e-01 -5.46830177e-01 5.72955795e-02
5.31415701e-01 2.32651547e-01 2.00196862e-01 -2.04798862e-01
-1.01310992e+00 8.34841251e-01 4.29444790e-01 2.45703802e-01
-5.07904708e-01 3.93588752e-01 4.02712703e-01 2.59482056e-01
7.11072922e-01 4.62390892e-02 -8.58571649e-01 -4.87605333e-01
-1.07568467e+00 2.43036866e-01 6.75322056e-01 7.60203898e-01
1.01184666e+00 7.97760189e-01 7.77986169e-01 5.53444088e-01
7.30172992e-01 1.14138603e+00 8.22832465e-01 -7.35987842e-01
1.65979832e-01 2.00574934e-01 1.59606069e-01 -1.06614292e+00
-9.51080799e-01 -6.82758510e-01 -1.25646448e+00 3.06154728e-01
-2.46915519e-01 -8.98194313e-01 -7.62325704e-01 1.35707295e+00
3.73036265e-01 6.17075503e-01 3.19237530e-01 7.18090355e-01
9.83476937e-01 5.33093929e-01 -2.21035480e-01 1.70324221e-01
1.30066764e+00 -8.91924918e-01 -9.53904986e-01 -7.98109889e-01
8.53153825e-01 -6.05489254e-01 5.67564607e-01 1.73544407e-01
-5.80610812e-01 -1.00524998e+00 -1.74219406e+00 -3.39838922e-01
-7.84214973e-01 2.87765563e-01 7.79665768e-01 1.03618705e+00
-1.13415194e+00 5.73572397e-01 -1.57807100e+00 -4.95997071e-01
-3.86604145e-02 5.35832405e-01 -1.39740154e-01 3.35842341e-01
-1.52968383e+00 1.02691984e+00 1.21934220e-01 4.82385218e-01
-4.77428943e-01 -2.76116997e-01 -1.37876403e+00 -3.15140367e-01
-2.37063646e-01 -1.12808609e+00 1.46019804e+00 -4.68817145e-01
-1.71988297e+00 5.93172550e-01 -3.28987330e-01 -1.10314703e+00
3.19224983e-01 -9.77249324e-01 -6.21537209e-01 -6.83698833e-01
6.45007286e-03 4.80458617e-01 7.04662979e-01 -7.70148754e-01
-9.72771108e-01 -3.59753489e-01 -9.80657246e-03 1.40261903e-01
-1.89991370e-01 -3.31876904e-01 -3.32512707e-02 -3.55148226e-01
2.79609293e-01 -1.11723614e+00 -3.46699834e-01 -3.96764487e-01
3.76815908e-02 1.89502031e-01 7.82376468e-01 -5.50146818e-01
1.17001772e+00 -2.18244410e+00 -6.81214780e-02 -1.72936469e-01
9.68759879e-02 9.14487764e-02 3.66599441e-01 2.56154418e-01
2.63727725e-01 -3.18426788e-01 4.35293578e-02 -9.05509889e-01
-1.90759171e-02 5.95886648e-01 -3.98189664e-01 7.36478686e-01
-2.49870777e-01 9.35549319e-01 -8.53543222e-01 2.50589717e-02
8.25289071e-01 7.98285902e-01 -4.09502476e-01 -1.03612155e-01
3.06997001e-01 6.72555506e-01 -3.99059564e-01 6.95562780e-01
6.86471939e-01 1.61571220e-01 -2.95300782e-01 -4.64361370e-01
-3.38333756e-01 4.91561413e-01 -1.09933174e+00 1.79113364e+00
-8.32510889e-01 8.37061286e-01 1.90618828e-01 -8.75769556e-01
9.27286029e-01 4.23793793e-02 6.97390378e-01 -6.71929121e-01
4.28630114e-01 4.49692249e-01 -3.37377578e-01 -3.50269824e-01
1.04157758e+00 -1.24628276e-01 -3.78418744e-01 7.71803856e-02
-4.25044037e-02 -3.09186399e-01 -1.72776580e-01 -2.82715738e-01
5.38415909e-01 2.49159783e-01 5.61785996e-01 -2.61932135e-01
6.14395320e-01 -3.76987427e-01 6.87163532e-01 7.19389141e-01
-2.78239876e-01 2.83430129e-01 -5.33835113e-01 -7.88483560e-01
-7.22087264e-01 -7.49581695e-01 -3.10694724e-01 1.20232987e+00
2.88250923e-01 -3.28732789e-01 -5.12335658e-01 1.44251168e-01
2.75691688e-01 1.14767917e-01 -5.01262903e-01 -7.90177435e-02
-5.26767731e-01 -9.94815230e-01 5.77354193e-01 9.41859603e-01
8.34753513e-01 -5.09157419e-01 -8.08135748e-01 2.21682474e-01
-2.53862552e-02 -1.01098275e+00 2.33769447e-01 2.71045029e-01
-1.42269933e+00 -6.85508072e-01 -2.49234065e-01 -6.20739102e-01
1.39108300e-01 4.29418534e-01 1.10794473e+00 -1.37232587e-01
4.69217688e-01 4.25548047e-01 -2.37757549e-01 -8.73503387e-01
3.01916525e-02 3.66320908e-01 9.36237991e-01 -2.71790177e-01
7.49707043e-01 -4.71765727e-01 -4.30087686e-01 7.74681568e-02
-5.40523350e-01 -6.85242936e-02 5.55900395e-01 6.56794250e-01
7.20585406e-01 -1.42696753e-01 2.73877650e-01 -3.22515100e-01
6.13386512e-01 -5.58950603e-01 -8.33612323e-01 -3.47151965e-01
-7.02915490e-01 -1.80604272e-02 9.84693617e-02 1.24634258e-01
-6.43065512e-01 3.30555856e-01 -8.38057935e-01 1.14637829e-01
-9.01149511e-02 1.00062048e+00 -6.35191146e-03 -3.66820186e-01
7.15892673e-01 6.66683614e-02 1.37176991e-01 -4.97216135e-01
2.78616935e-01 8.74096274e-01 6.98976338e-01 -9.60884690e-02
5.97794652e-01 5.35676003e-01 -6.02977052e-02 -1.27954221e+00
-7.53817677e-01 -3.08794498e-01 -5.21793902e-01 -3.02724183e-01
6.42555773e-01 -1.51925910e+00 -4.83592629e-01 7.12021410e-01
-8.98230195e-01 -2.83798486e-01 -1.97926238e-01 8.84096622e-01
-4.52704608e-01 2.02288434e-01 -7.20623851e-01 -7.76744902e-01
-3.99355620e-01 -1.40830791e+00 1.58172905e+00 6.65498376e-01
-2.00377358e-03 -1.38388169e+00 5.55571496e-01 1.56264588e-01
7.56772399e-01 3.29055935e-01 -3.80678385e-01 -2.48669028e-01
-4.75073665e-01 -7.04226792e-01 2.53080875e-01 4.76927936e-01
2.14694411e-01 -2.20307067e-01 -1.17124474e+00 -4.64356244e-01
3.69918972e-01 -6.37278110e-02 8.60868573e-01 8.83502781e-01
4.38922286e-01 2.59948105e-01 -5.59252143e-01 1.07525456e+00
1.26795518e+00 8.34329128e-02 4.74646181e-01 5.92751741e-01
6.63489401e-01 -6.30838126e-02 5.95819592e-01 2.91738480e-01
6.93120241e-01 4.21847790e-01 5.08635342e-01 -3.55042279e-01
1.98195308e-01 -7.88957924e-02 5.57751000e-01 1.26470637e+00
-3.43725055e-01 7.44587332e-02 -8.37188184e-01 1.77208930e-01
-1.85927129e+00 -5.36422908e-01 -1.25089377e-01 1.92736387e+00
3.08380872e-01 1.54309168e-01 -2.45800689e-01 2.05294535e-01
3.24058324e-01 3.65748376e-01 -5.60097992e-01 -2.21176267e-01
-2.83617437e-01 -7.19604120e-02 8.71883750e-01 9.23660100e-01
-1.62348545e+00 9.27452981e-01 7.01447105e+00 2.35197008e-01
-1.40985417e+00 3.24076831e-01 1.77226216e-01 1.96951360e-01
9.46649387e-02 -4.99952346e-01 -1.24198115e+00 4.66645658e-01
1.29616189e+00 1.59122553e-02 1.12698160e-01 1.19070566e+00
4.12051886e-01 -4.41102564e-01 -7.53976226e-01 1.24089956e+00
4.88572940e-02 -1.28975618e+00 -5.88767231e-01 2.48648778e-01
8.86891603e-01 1.07272983e+00 -3.10996342e-02 4.07095760e-01
7.17818141e-02 -5.61156034e-01 5.95660865e-01 4.93443847e-01
5.47732174e-01 -5.89466333e-01 1.47638643e+00 1.12163916e-01
-1.51548338e+00 -1.00284606e-01 -5.48097372e-01 -9.93871629e-01
5.36553979e-01 1.03019845e+00 -2.80849189e-01 5.85968971e-01
8.48267913e-01 1.23764205e+00 -3.24683189e-01 9.32452500e-01
-2.93546140e-01 4.09160942e-01 -7.55135596e-01 -1.07499585e-01
1.28700018e-01 -3.73755068e-01 2.36480907e-01 1.17763388e+00
7.19994485e-01 -1.58722207e-01 2.48509601e-01 -9.25313532e-02
8.14501494e-02 -4.04064953e-01 -1.05789709e+00 2.98358291e-01
4.83025491e-01 9.56458986e-01 -3.55973542e-01 -5.27734578e-01
-8.22842419e-01 4.71343666e-01 -3.78657788e-01 2.91621655e-01
-8.46784294e-01 -3.81652921e-01 1.31998682e+00 -2.47454435e-01
1.04884528e-01 -1.10414243e+00 -6.72932863e-01 -1.32090163e+00
-2.92341828e-01 -4.31853920e-01 -7.42646158e-02 -6.49300396e-01
-8.40111792e-01 5.33002555e-01 -2.19206423e-01 -1.50420892e+00
-6.02443814e-01 -8.95885110e-01 -1.60818815e-01 6.63318276e-01
-1.56495512e+00 -8.38157415e-01 -5.88758707e-01 2.99657136e-01
3.83076102e-01 -3.05953830e-01 7.81076550e-01 7.01012671e-01
-3.90786469e-01 2.47292385e-01 6.08467042e-01 1.93708003e-01
5.86185575e-01 -9.27954912e-01 9.89739776e-01 1.17651188e+00
1.71989456e-01 6.37766361e-01 9.73427534e-01 -5.47447741e-01
-1.66323602e+00 -9.44296122e-01 8.59490216e-01 -5.27269185e-01
6.52851224e-01 -4.12306011e-01 -4.08332586e-01 9.07051027e-01
1.22895978e-01 -3.65111455e-02 7.90136099e-01 1.69851437e-01
1.85015485e-01 -6.09873235e-01 -8.59566927e-01 2.90823340e-01
7.66788900e-01 -6.58386469e-01 -5.89181900e-01 1.82374731e-01
4.82020676e-01 -1.12320125e+00 -6.73897028e-01 8.16489697e-01
1.00385964e+00 -1.08154690e+00 1.10020363e+00 -1.91060796e-01
-1.07946433e-01 -5.66415131e-01 -2.79723346e-01 -1.27154183e+00
-2.48792022e-01 -3.68251562e-01 -6.66179836e-01 5.22812665e-01
1.12680525e-01 -1.02941537e+00 9.19009268e-01 3.11740607e-01
-5.24636209e-01 -5.59310257e-01 -1.01657999e+00 -4.42974716e-01
-1.17372274e-01 -9.28330243e-01 6.29373074e-01 3.90082926e-01
-4.03634459e-01 4.12462562e-01 -6.32535696e-01 4.01824862e-01
4.53307480e-01 -1.24038205e-01 1.02538192e+00 -1.33166814e+00
1.79763243e-01 -1.56037658e-01 -9.13349807e-01 -1.63048100e+00
6.59456849e-02 -2.09262401e-01 2.46443167e-01 -1.52087808e+00
-7.40734637e-01 -1.69379830e-01 -2.85434455e-01 7.40930215e-02
-7.32113272e-02 6.66939795e-01 -3.29702497e-01 -3.09172249e-03
-4.16824192e-01 5.74151754e-01 6.44767523e-01 -1.90367043e-01
-2.54848748e-01 3.92472029e-01 -5.44050336e-01 8.88073921e-01
1.09163034e+00 -1.81540810e-02 -3.07284594e-01 -9.30944383e-01
8.86419713e-01 -4.33977097e-01 1.18282497e-01 -1.68099892e+00
4.14695650e-01 4.32011783e-01 5.44714987e-01 -8.13165247e-01
2.98375428e-01 -8.09768856e-01 7.96196908e-02 7.31453001e-01
5.22017777e-01 4.65865135e-01 4.60999399e-01 5.98319948e-01
-5.61370194e-01 3.40506524e-01 5.85011840e-01 2.42075756e-01
-1.15648723e+00 1.64474383e-01 -8.02897334e-01 -4.89106655e-01
4.54482436e-01 -4.93555516e-01 1.23065017e-01 -6.59595847e-01
-5.75595498e-01 2.60494262e-01 3.39603871e-01 5.82897305e-01
5.05090058e-01 -1.43392467e+00 -3.05012494e-01 6.10249102e-01
-1.42203867e-01 1.25546917e-01 2.87767559e-01 9.48811769e-01
-7.47719228e-01 7.55186498e-01 -1.67591289e-01 -8.13618600e-01
-8.30875754e-01 1.69663012e-01 6.34481966e-01 1.17898174e-01
-3.41764748e-01 8.36376607e-01 -3.39236945e-01 -6.32873297e-01
2.64813483e-01 -1.02240574e+00 -2.02157289e-01 -5.53416386e-02
6.49651825e-01 5.11545181e-01 5.25716782e-01 -8.78412008e-01
-5.11239111e-01 9.44957793e-01 5.17299235e-01 -5.25237527e-03
9.77235436e-01 -5.88146389e-01 -1.13062328e-02 6.20668888e-01
1.02759004e+00 -1.41453007e-02 -1.19949889e+00 6.34780154e-02
-3.48613150e-02 1.03071570e-01 3.84097159e-01 -2.81163216e-01
-1.07026613e+00 8.88532877e-01 1.17089140e+00 1.38808891e-01
1.16270351e+00 -3.79866838e-01 1.13530290e+00 7.16031253e-01
8.64055336e-01 -1.05662286e+00 -5.28812528e-01 1.22917151e+00
3.64091605e-01 -1.42785108e+00 2.57780164e-01 1.84591517e-01
1.04568146e-01 9.80570674e-01 2.46908322e-01 -3.30373079e-01
1.27817297e+00 7.64770567e-01 5.06939948e-01 1.45051301e-01
-8.30360875e-03 -3.30409229e-01 1.30960420e-01 6.26194715e-01
8.13471198e-01 1.64332986e-01 -3.55312109e-01 4.74672377e-01
-7.48521805e-01 2.46388897e-01 2.59535611e-01 1.17064059e+00
-7.46185720e-01 -1.00471759e+00 -4.98579919e-01 4.43247288e-01
-2.69073755e-01 3.83007303e-02 4.04566713e-02 4.76049751e-01
3.30302685e-01 1.00055635e+00 2.72491187e-01 -9.41556215e-01
1.85012937e-01 -4.55511175e-02 2.52493620e-01 -1.90878481e-01
-1.73106000e-01 -1.55252712e-02 1.44314080e-01 -6.15790069e-01
-8.59723747e-01 -5.70107162e-01 -1.09129739e+00 -4.84643549e-01
-2.84247130e-01 7.20521137e-02 1.27037811e+00 1.18835366e+00
6.52544379e-01 5.48551321e-01 2.17202276e-01 -1.65849686e+00
-1.93311557e-01 -1.10615873e+00 -4.13586915e-01 -4.93498117e-01
8.95063996e-01 -9.45831954e-01 -2.45985836e-01 8.66602510e-02] | [7.459315299987793, -1.9654712677001953] |
31234063-7e8d-4105-8496-270c7b8f33f9 | it-is-all-connected-a-new-graph-formulation | 2303.13177 | null | https://arxiv.org/abs/2303.13177v1 | https://arxiv.org/pdf/2303.13177v1.pdf | It is all Connected: A New Graph Formulation for Spatio-Temporal Forecasting | With an ever-increasing number of sensors in modern society, spatio-temporal time series forecasting has become a de facto tool to make informed decisions about the future. Most spatio-temporal forecasting models typically comprise distinct components that learn spatial and temporal dependencies. A common methodology employs some Graph Neural Network (GNN) to capture relations between spatial locations, while another network, such as a recurrent neural network (RNN), learns temporal correlations. By representing every recorded sample as its own node in a graph, rather than all measurements for a particular location as a single node, temporal and spatial information is encoded in a similar manner. In this setting, GNNs can now directly learn both temporal and spatial dependencies, jointly, while also alleviating the need for additional temporal networks. Furthermore, the framework does not require aligned measurements along the temporal dimension, meaning that it also naturally facilitates irregular time series, different sampling frequencies or missing data, without the need for data imputation. To evaluate the proposed methodology, we consider wind speed forecasting as a case study, where our proposed framework outperformed other spatio-temporal models using GNNs with either Transformer or LSTM networks as temporal update functions. | ['Paal Engelstad', 'Roy Stenbro', 'Narada Dilp Warakagoda', 'Lars Ødegaard Bentsen'] | 2023-03-23 | null | null | null | null | ['irregular-time-series', 'spatio-temporal-forecasting'] | ['time-series', 'time-series'] | [ 4.80739353e-03 -1.11627869e-01 -2.02954888e-01 -2.59428561e-01
9.02808979e-02 -4.95290488e-01 9.20835078e-01 4.94645536e-01
-2.85370648e-01 7.84015238e-01 3.11520278e-01 -5.88715434e-01
-4.96776253e-01 -1.25499940e+00 -7.29803681e-01 -7.23583877e-01
-6.38122499e-01 1.25289723e-01 9.48994532e-02 -1.58683553e-01
-6.49416596e-02 5.42006433e-01 -1.25572097e+00 -3.46184433e-01
8.94129753e-01 1.16976452e+00 1.97818458e-01 6.70004010e-01
-1.70750514e-01 1.11661398e+00 -4.49724495e-01 -2.61789024e-01
1.17935650e-01 -1.70078307e-01 -3.94098312e-01 -1.43704921e-01
-1.93935726e-02 -1.08669445e-01 -6.42686725e-01 6.51167393e-01
1.33876175e-01 6.08156621e-01 3.51002783e-01 -1.24916434e+00
-6.80146098e-01 7.13122070e-01 -5.01156747e-01 2.90919989e-01
-1.39186516e-01 -6.88607171e-02 7.87812293e-01 -2.85147130e-01
2.15651482e-01 8.46412003e-01 8.35111916e-01 -1.30501524e-01
-1.22517526e+00 -4.99488562e-01 5.35305083e-01 2.07433149e-01
-1.22554123e+00 -2.83130348e-01 1.14458835e+00 -4.24639493e-01
9.37813103e-01 1.02178201e-01 8.93440783e-01 9.61459041e-01
4.94966537e-01 3.07856739e-01 5.71339607e-01 -1.78985472e-03
2.04808041e-01 -2.75762528e-01 -1.07213281e-01 1.74052566e-01
-1.33535191e-01 1.63469508e-01 -4.44694966e-01 -1.71617538e-01
7.21181989e-01 7.01211691e-01 -1.69867843e-01 -2.21160114e-01
-1.13281572e+00 6.10630631e-01 8.71753514e-01 6.73036993e-01
-8.06034088e-01 4.57180530e-01 3.75315070e-01 1.86143041e-01
9.98822272e-01 -2.30126232e-02 -3.61725867e-01 -9.56512615e-02
-1.14022875e+00 -8.28548521e-02 5.05698621e-01 4.94145721e-01
7.92194366e-01 5.23356080e-01 1.68533981e-01 5.17124772e-01
1.16681501e-01 7.06114292e-01 5.35689414e-01 -5.35129786e-01
7.16848731e-01 5.57338655e-01 2.20106676e-01 -1.71494460e+00
-7.09014177e-01 -8.31606746e-01 -1.67752862e+00 -2.09772334e-01
2.27167830e-01 -5.23292780e-01 -8.34607661e-01 2.00475645e+00
2.72003889e-01 9.97153699e-01 -6.68520778e-02 6.07029557e-01
2.96008795e-01 1.04500568e+00 8.27865824e-02 -3.24859679e-01
8.48056197e-01 -6.01211548e-01 -9.24138188e-01 -1.52965963e-01
6.19604647e-01 -2.01176822e-01 3.22163910e-01 4.83251959e-02
-5.85110724e-01 -4.81334299e-01 -7.86448896e-01 1.76164210e-01
-8.37766469e-01 2.75363717e-02 6.73674762e-01 1.61261022e-01
-1.31224501e+00 8.24337780e-01 -1.04366350e+00 -4.32687014e-01
-1.47666797e-01 1.72223538e-01 -2.79466629e-01 3.45632970e-01
-1.52510536e+00 7.26409316e-01 4.33285803e-01 7.06436813e-01
-2.99084157e-01 -6.17104530e-01 -8.92879188e-01 2.25145116e-01
9.54148993e-02 -5.25957286e-01 7.03094900e-01 -7.96068072e-01
-1.31574249e+00 -9.40972641e-02 -5.54248929e-01 -8.79806101e-01
1.98473155e-01 9.31018889e-02 -9.49244797e-01 -3.29337031e-01
-6.78440854e-02 1.74965546e-01 9.76367235e-01 -8.65885735e-01
-6.81873918e-01 -4.46921617e-01 8.02569017e-02 -2.42796186e-02
-5.08150220e-01 -5.58562040e-01 2.00938694e-02 -8.77979040e-01
1.67296022e-01 -8.41757178e-01 -4.41350192e-01 -3.54373157e-01
-3.30017418e-01 -2.13338464e-01 9.46582317e-01 -9.46945727e-01
1.51700711e+00 -1.98611200e+00 2.49381334e-01 4.44537014e-01
2.33261168e-01 1.11280456e-01 -1.37626320e-01 8.59800935e-01
-3.78008038e-02 -2.65180916e-02 -4.79954869e-01 -6.56331599e-01
-2.29556456e-01 5.12237906e-01 -6.57119989e-01 4.78293777e-01
2.07811762e-02 9.01708066e-01 -1.03753412e+00 1.03405274e-01
5.74130774e-01 8.07912707e-01 3.33854295e-02 1.52839487e-02
-2.38062128e-01 7.89507210e-01 -2.89225757e-01 -9.20105353e-02
5.40458798e-01 -4.69280928e-01 3.34202081e-01 1.09613381e-01
-4.37025696e-01 2.93346137e-01 -1.16757059e+00 1.56288648e+00
-9.92644489e-01 7.34055340e-01 -2.81923681e-01 -1.29485214e+00
1.09100890e+00 5.95365465e-01 6.68990672e-01 -1.02818000e+00
-2.73785889e-01 2.77108829e-02 -2.30422199e-01 -4.16389018e-01
4.85845089e-01 2.39509329e-01 1.18477911e-01 4.90491837e-01
-2.27667615e-01 4.79101092e-01 -5.46810254e-02 -1.25076577e-01
1.00922942e+00 -1.67770647e-02 4.41526398e-02 1.29816607e-01
4.56600696e-01 -1.71837732e-01 6.83696449e-01 5.21880507e-01
3.50383699e-01 1.17677011e-01 2.26034045e-01 -1.02288151e+00
-8.89175117e-01 -6.85976923e-01 1.62613258e-01 8.34520578e-01
-1.04413763e-01 -2.04073772e-01 -1.17929764e-01 -2.03947455e-01
6.99902400e-02 7.99279749e-01 -8.37515473e-01 1.37920931e-01
-5.86419046e-01 -7.63017416e-01 2.55531579e-01 4.13735420e-01
3.78554463e-01 -1.06319678e+00 -6.51404500e-01 5.70873976e-01
-2.49739796e-01 -9.90225971e-01 -1.10087149e-01 2.72677481e-01
-1.15923822e+00 -7.07663715e-01 -7.17588603e-01 -3.07730645e-01
4.66515571e-01 4.32078660e-01 9.66493249e-01 -9.37936530e-02
4.34675455e-01 3.78920548e-02 -3.87520671e-01 -1.44683614e-01
5.87020814e-02 4.34160471e-01 1.01207662e-02 4.84286815e-01
1.25481918e-01 -1.50802028e+00 -6.42267764e-01 -5.24806082e-02
-9.91877794e-01 1.89031705e-01 2.99231559e-01 6.42934740e-01
2.14399293e-01 2.54485130e-01 8.53607595e-01 -7.08887935e-01
4.06223834e-01 -1.03890204e+00 -7.44106770e-01 2.57842720e-01
-4.59700376e-01 -2.41136998e-02 1.18219173e+00 -4.29027915e-01
-9.17515993e-01 1.09796077e-02 1.08207911e-01 -4.14030701e-01
1.05841225e-02 1.18004465e+00 3.31366003e-01 2.02407643e-01
2.81469047e-01 6.18502080e-01 -3.28513354e-01 -5.59238195e-01
4.54661459e-01 4.02000487e-01 4.70380425e-01 -7.41286650e-02
8.99913967e-01 5.13505459e-01 2.45693445e-01 -9.59010899e-01
-5.22212982e-01 -3.60603064e-01 -7.90679038e-01 -2.67902583e-01
6.73649311e-01 -8.42010498e-01 -8.84245932e-01 4.15088058e-01
-1.30543244e+00 -4.63872284e-01 -1.93466187e-01 6.38766706e-01
-2.16924530e-02 2.01984625e-02 -3.92023236e-01 -1.09010112e+00
-1.60181686e-01 -5.69425821e-01 8.92794788e-01 2.83499733e-02
2.43556462e-02 -1.69372582e+00 1.75320864e-01 -2.72511959e-01
7.50967979e-01 6.52389944e-01 8.50728095e-01 -3.10352921e-01
-5.92321277e-01 -3.09390545e-01 -2.37460867e-01 -4.30506356e-02
5.46844959e-01 -1.18008018e-01 -7.87683189e-01 -2.08792701e-01
8.57410729e-02 3.83385926e-01 6.93164170e-01 6.40762448e-01
1.04073620e+00 -7.00696290e-01 -1.98498577e-01 7.38967657e-01
1.31201673e+00 4.08296496e-01 3.38705212e-01 1.15317494e-01
1.04446077e+00 6.90710723e-01 -1.14988022e-01 5.86564124e-01
9.18604374e-01 5.70293367e-01 5.47223151e-01 -2.41836399e-01
2.66408175e-01 -3.40701699e-01 1.08495913e-03 1.11381865e+00
-3.41266841e-01 -6.46241426e-01 -1.08314002e+00 8.31447601e-01
-2.20960188e+00 -1.05598915e+00 -1.66694865e-01 2.28154683e+00
1.66192994e-01 -2.33534202e-01 1.37263522e-01 3.31913799e-01
5.83458483e-01 8.15876305e-01 -7.54412651e-01 -7.30445459e-02
-1.77959085e-01 -1.00764325e-02 8.29300046e-01 4.96834487e-01
-9.37114000e-01 5.23095191e-01 5.33135843e+00 4.71014678e-01
-1.59555268e+00 8.78370702e-02 6.42693579e-01 1.75336182e-01
-4.56765831e-01 -2.37815790e-02 -1.37422994e-01 5.39770067e-01
1.27524412e+00 -1.77772164e-01 6.89910650e-01 4.53393400e-01
7.66843140e-01 5.54171503e-02 -6.41057670e-01 7.50649095e-01
-2.73000062e-01 -1.44345558e+00 -1.68991287e-03 9.93777066e-02
6.10012531e-01 2.53347456e-01 1.52888820e-01 -2.71667745e-02
4.34565604e-01 -1.03609776e+00 3.82662714e-01 9.71194506e-01
4.30224121e-01 -8.02219272e-01 7.50388324e-01 5.74027896e-01
-1.65282381e+00 -1.26158252e-01 -1.42250165e-01 -4.17165279e-01
4.33026195e-01 9.05752718e-01 -4.20197636e-01 9.74262059e-01
6.65258527e-01 1.36182964e+00 -3.36940408e-01 8.31887066e-01
-2.64009293e-02 7.41338670e-01 -5.62661409e-01 8.18928182e-02
3.41743499e-01 -6.54483140e-01 4.84060973e-01 7.55208611e-01
7.41352618e-01 -6.55270368e-02 9.71329361e-02 3.42238218e-01
-5.99072361e-03 -1.57523364e-01 -8.43513250e-01 7.83276036e-02
5.53957283e-01 9.88894165e-01 -5.86346269e-01 -2.42685869e-01
-5.10271788e-01 6.54184699e-01 4.19764966e-01 8.19295824e-01
-6.90698147e-01 -1.24273524e-01 5.45214176e-01 1.90818846e-01
1.87989503e-01 -6.14324152e-01 -1.93359986e-01 -1.01423562e+00
2.32407942e-01 -3.04453135e-01 2.74462491e-01 -7.13832021e-01
-1.26364183e+00 7.19810724e-01 -1.67832017e-01 -1.43889487e+00
-6.05971098e-01 -6.96270540e-02 -6.46666169e-01 1.00791121e+00
-1.63094270e+00 -1.37622738e+00 -1.97673962e-01 5.58280945e-01
1.23419456e-01 1.65107012e-01 6.80647731e-01 3.86277407e-01
-6.81257069e-01 5.17368950e-02 4.37264264e-01 1.12218991e-01
1.96633413e-01 -1.15023160e+00 8.63699436e-01 1.06345034e+00
3.56618941e-01 5.55763841e-01 4.94746119e-01 -6.51357293e-01
-1.33881176e+00 -1.58829713e+00 1.27879071e+00 4.47777100e-02
1.05873215e+00 -3.05605918e-01 -1.13681269e+00 8.98151994e-01
5.83680756e-02 2.49396950e-01 1.82721093e-01 3.38264555e-01
-3.69138271e-02 -5.11535525e-01 -6.41041517e-01 4.44237530e-01
9.73179042e-01 -6.67258501e-01 -6.35249913e-02 4.06542689e-01
8.31448853e-01 -2.19428986e-01 -9.77353632e-01 3.75037640e-01
3.40213269e-01 -9.66779292e-01 7.33089507e-01 -1.98578715e-01
1.26158655e-01 -4.32145059e-01 -9.70749534e-04 -1.66094327e+00
-4.77055579e-01 -5.60176075e-01 -4.34080660e-01 1.19580400e+00
4.01427865e-01 -1.04602289e+00 7.49694586e-01 4.95008707e-01
9.96365175e-02 -5.42175233e-01 -1.17908394e+00 -6.98960543e-01
-2.07958937e-01 -6.73907518e-01 1.13371384e+00 1.18445957e+00
-3.61694545e-01 2.05612451e-01 -9.98083830e-01 5.51228285e-01
3.64728630e-01 1.68800920e-01 5.80476880e-01 -1.45819998e+00
-1.27057627e-01 -2.07604945e-01 -4.12039429e-01 -1.17004836e+00
1.96524411e-01 -5.53347468e-01 -2.84874231e-01 -1.56826580e+00
-4.34404910e-01 -6.18296623e-01 -5.00894427e-01 5.00666678e-01
1.18167818e-01 -1.15796430e-02 -5.36281094e-02 1.00696385e-01
-1.08391985e-01 8.54368806e-01 8.27390254e-01 -1.33652031e-01
-5.14949083e-01 3.03251237e-01 2.93128286e-02 5.06533861e-01
7.98614740e-01 -4.14831907e-01 -8.02584052e-01 -7.80006170e-01
5.95714748e-01 7.20105469e-01 4.50643688e-01 -1.12479591e+00
6.86334670e-01 -2.03945130e-01 4.17070359e-01 -7.43301272e-01
2.74830520e-01 -1.14780700e+00 8.38981509e-01 2.08843037e-01
-8.60514417e-02 6.32166862e-01 2.29808122e-01 8.25811267e-01
-4.06231701e-01 4.20694262e-01 9.09933541e-03 3.05057794e-01
-5.78713834e-01 6.46180391e-01 -5.20140290e-01 -4.49902475e-01
7.12891817e-01 -2.42133543e-01 -2.50861824e-01 -7.65280724e-01
-7.48624444e-01 4.29133594e-01 2.01107800e-01 4.90354002e-01
2.39477515e-01 -1.30474651e+00 -4.05066222e-01 4.23972048e-02
-1.71717599e-01 1.09236717e-01 7.06231713e-01 8.55249226e-01
-8.32939148e-02 5.16303957e-01 1.89557612e-01 -6.05944812e-01
-4.64816958e-01 5.79793334e-01 3.58293384e-01 -4.80869234e-01
-6.59569204e-01 3.59963238e-01 -1.82274789e-01 -5.19058883e-01
-4.31653708e-02 -5.74762285e-01 -4.28115219e-01 2.87675232e-01
2.85435319e-01 3.56432348e-01 1.71582520e-01 -7.72242606e-01
-1.90191701e-01 4.79550093e-01 6.66610420e-01 -1.34086043e-01
1.69354463e+00 -4.26493973e-01 -2.09547818e-01 1.12259734e+00
9.77366388e-01 -2.67980665e-01 -1.14932501e+00 -4.96386260e-01
1.56781912e-01 -1.25247851e-01 3.18627238e-01 -4.36027527e-01
-1.39458716e+00 8.35692048e-01 3.30020756e-01 7.98700631e-01
1.30221891e+00 -6.32523477e-01 1.03802431e+00 2.64581144e-01
3.68773222e-01 -6.21911108e-01 -5.38077056e-01 5.65743387e-01
7.12125361e-01 -1.02025282e+00 -1.98283762e-01 -2.57596038e-02
-3.30870807e-01 1.03420818e+00 -1.59319304e-02 -1.84438705e-01
1.05558038e+00 -2.87644863e-02 -1.74251258e-01 -4.69481945e-02
-7.32289612e-01 -2.07074985e-01 4.09760058e-01 5.15426099e-01
4.02228713e-01 2.29805216e-01 1.99329183e-01 1.19341120e-01
-2.73519397e-01 6.61696792e-02 1.69095039e-01 5.23034751e-01
1.63071260e-01 -9.03503358e-01 -4.78959531e-02 6.45210683e-01
-1.15808636e-01 -1.86378639e-02 2.71603465e-01 7.13831604e-01
4.00499478e-02 1.20488846e+00 4.83349353e-01 -5.75076401e-01
2.96244144e-01 -1.32727087e-01 -2.00031281e-01 -2.03912035e-01
-5.24524510e-01 -2.04376683e-01 -1.53250456e-01 -5.15646994e-01
-7.05075145e-01 -5.12849689e-01 -9.08398151e-01 -6.39526486e-01
-1.13419048e-03 2.77452469e-01 6.90321147e-01 1.11657810e+00
6.66320860e-01 7.17799664e-01 8.88344049e-01 -1.02292824e+00
2.38678113e-01 -1.00200665e+00 -6.25418782e-01 7.68936798e-02
8.35867405e-01 -4.71578747e-01 -1.78637967e-01 -3.08623407e-02] | [6.63401985168457, 2.748762369155884] |
00a29c3f-8a18-49f7-87b4-b4ecf1f2bcca | segnerf-3d-part-segmentation-with-neural | 2211.11215 | null | https://arxiv.org/abs/2211.11215v2 | https://arxiv.org/pdf/2211.11215v2.pdf | SegNeRF: 3D Part Segmentation with Neural Radiance Fields | Recent advances in Neural Radiance Fields (NeRF) boast impressive performances for generative tasks such as novel view synthesis and 3D reconstruction. Methods based on neural radiance fields are able to represent the 3D world implicitly by relying exclusively on posed images. Yet, they have seldom been explored in the realm of discriminative tasks such as 3D part segmentation. In this work, we attempt to bridge that gap by proposing SegNeRF: a neural field representation that integrates a semantic field along with the usual radiance field. SegNeRF inherits from previous works the ability to perform novel view synthesis and 3D reconstruction, and enables 3D part segmentation from a few images. Our extensive experiments on PartNet show that SegNeRF is capable of simultaneously predicting geometry, appearance, and semantic information from posed images, even for unseen objects. The predicted semantic fields allow SegNeRF to achieve an average mIoU of $\textbf{30.30%}$ for 2D novel view segmentation, and $\textbf{37.46%}$ for 3D part segmentation, boasting competitive performance against point-based methods by using only a few posed images. Additionally, SegNeRF is able to generate an explicit 3D model from a single image of an object taken in the wild, with its corresponding part segmentation. | ['Bernard Ghanem', 'Silvio Giancola', 'Sara Rojas', 'Jesus Zarzar'] | 2022-11-21 | null | null | null | null | ['3d-part-segmentation'] | ['computer-vision'] | [ 3.73756438e-01 4.06269729e-01 2.44142056e-01 -5.54969251e-01
-7.05548584e-01 -5.26348531e-01 6.41471267e-01 -3.04636598e-01
-4.03571948e-02 3.89942795e-01 -3.11251104e-01 -5.90811782e-02
8.96812975e-03 -1.16188097e+00 -1.19886625e+00 -4.70805466e-01
2.85000861e-01 6.17601812e-01 2.38062903e-01 -3.93006891e-01
-1.18185192e-01 8.33472550e-01 -1.78616846e+00 4.60462831e-02
8.10921192e-01 1.15677524e+00 4.35721010e-01 3.91296566e-01
-2.14696839e-01 3.49231750e-01 -4.61866647e-01 -5.32865644e-01
5.64160585e-01 -3.52044314e-01 -7.21707523e-01 3.81560951e-01
8.94596338e-01 -4.50146943e-01 -3.06285322e-01 7.92269766e-01
3.64553183e-01 2.16256112e-01 7.57753134e-01 -1.02486181e+00
-4.08778965e-01 3.92930835e-01 -4.18204129e-01 -1.66699603e-01
3.63210976e-01 1.64135143e-01 9.70224738e-01 -8.96727145e-01
7.67644227e-01 1.13912249e+00 6.74294531e-01 4.83251363e-01
-1.36689961e+00 -4.67573792e-01 2.09511504e-01 -2.17727885e-01
-1.28588820e+00 -2.61658490e-01 1.11254978e+00 -2.35982299e-01
1.15495741e+00 3.00373942e-01 8.37754726e-01 1.02473414e+00
1.93669736e-01 1.07399213e+00 1.00399339e+00 -9.01640877e-02
2.80722737e-01 -6.50873855e-02 -3.70743811e-01 7.59552896e-01
-1.13733090e-01 1.61024988e-01 -4.03292537e-01 2.64881700e-01
1.01365089e+00 2.06761695e-02 -4.11627948e-01 -4.55918252e-01
-1.22306526e+00 6.76475406e-01 9.60525930e-01 1.56663164e-01
-4.50488597e-01 3.60845149e-01 -1.81715250e-01 2.28688885e-02
8.68679464e-01 4.73789901e-01 -5.38316011e-01 2.53381193e-01
-1.01756072e+00 4.83851701e-01 4.54434842e-01 1.16690838e+00
1.12353194e+00 3.51933986e-01 1.04269736e-01 8.27186823e-01
2.10191458e-01 8.84581208e-01 4.07774113e-02 -1.14134073e+00
2.97842056e-01 7.40433276e-01 -1.25051945e-01 -7.13370323e-01
-3.76612514e-01 -5.34069419e-01 -8.40499222e-01 2.79804736e-01
3.24329771e-02 1.55006245e-01 -1.43363607e+00 1.52817822e+00
4.42666590e-01 7.31686577e-02 -1.78496808e-01 8.39383721e-01
9.27353978e-01 7.17074990e-01 -2.65194029e-01 2.09140003e-01
1.08871996e+00 -8.65871847e-01 -1.25012184e-02 -5.48121929e-01
3.04606825e-01 -6.58335090e-01 7.90153921e-01 4.62894022e-01
-1.21694279e+00 -7.84217656e-01 -8.65983367e-01 -1.14123143e-01
-1.91997766e-01 -1.65271968e-01 8.57971728e-01 5.50750196e-01
-1.23365819e+00 5.87829828e-01 -6.57453656e-01 -1.69142941e-03
7.01581895e-01 3.35436106e-01 -5.44396162e-01 -3.62201840e-01
-8.27815473e-01 5.12222588e-01 3.86007905e-01 2.34561488e-01
-9.07621861e-01 -7.55261183e-01 -1.19796073e+00 9.10415947e-02
4.51068908e-01 -1.18710470e+00 1.19605207e+00 -8.73621583e-01
-1.32897878e+00 9.77487504e-01 1.04461707e-01 -4.25339013e-01
4.94198680e-01 -1.98849410e-01 -2.35986948e-01 3.89236569e-01
1.48287863e-01 1.23416305e+00 9.22931671e-01 -1.47109485e+00
-4.11127687e-01 -5.55401683e-01 3.08924884e-01 2.85970777e-01
3.13911736e-01 -5.52880108e-01 -3.23426962e-01 -8.01367700e-01
6.09244227e-01 -7.51790285e-01 -3.35616857e-01 9.95794609e-02
-4.47529107e-01 -1.46576121e-01 7.37646937e-01 -4.06266391e-01
2.52307385e-01 -1.95013762e+00 1.44457906e-01 8.99050087e-02
2.37153500e-01 7.33830780e-02 -7.35013932e-02 2.41130572e-02
-2.10310929e-02 4.13293801e-02 -4.73033607e-01 -3.64539742e-01
1.36652946e-01 1.22922063e-01 -1.80506587e-01 3.06523889e-01
2.76397347e-01 1.24715066e+00 -6.80636466e-01 -2.48248160e-01
5.53383291e-01 7.02020526e-01 -6.64992988e-01 9.03256387e-02
-7.30144322e-01 4.02061433e-01 -4.82435167e-01 7.63919294e-01
9.97860849e-01 -2.40762293e-01 -1.16102524e-01 -2.70592511e-01
3.77731360e-02 4.90947403e-02 -1.00365937e+00 2.08594608e+00
-6.19268239e-01 3.83726269e-01 9.84049961e-02 -1.01621234e+00
1.13640380e+00 1.75312608e-02 6.12994909e-01 -7.89331198e-01
2.58850455e-01 2.40087762e-01 -4.74349350e-01 -5.54053970e-02
5.03572822e-01 -4.58240598e-01 -2.43395448e-01 3.63145649e-01
4.28744793e-01 -8.63996983e-01 -1.02990165e-01 8.61822739e-02
7.20798552e-01 6.55116320e-01 -1.60429224e-01 -2.09764112e-02
2.97795534e-01 -5.05958498e-02 3.05944890e-01 6.49997175e-01
3.24347526e-01 9.67797995e-01 -6.08430430e-02 -4.96892720e-01
-9.54852700e-01 -1.50788188e+00 -1.72058523e-01 6.37903631e-01
5.52181721e-01 -8.97826478e-02 -8.50023448e-01 -8.73719275e-01
-8.03446304e-03 8.22709322e-01 -4.02029842e-01 -1.54207826e-01
-6.25960827e-01 -4.83219326e-01 2.39923850e-01 5.57235479e-01
8.11809897e-01 -9.95917261e-01 -8.13021839e-01 1.04677640e-01
-3.48996490e-01 -1.23085725e+00 -1.76074833e-01 2.73190141e-01
-1.11860979e+00 -8.91735375e-01 -1.03013194e+00 -7.20266461e-01
5.99355340e-01 3.45117956e-01 1.42860913e+00 -1.79265499e-01
-3.57100666e-01 3.76737535e-01 -3.03835630e-01 -2.73354620e-01
-2.90837377e-01 3.60449478e-02 -3.45532805e-01 3.32249037e-04
-1.19671682e-02 -8.95142078e-01 -7.52782524e-01 4.57847953e-01
-1.12782288e+00 4.47294503e-01 7.81770885e-01 6.69383705e-01
9.82905447e-01 -6.80036619e-02 2.33112514e-01 -9.08153534e-01
-3.41972768e-01 -2.81829506e-01 -4.90821540e-01 4.69147488e-02
-4.19731379e-01 9.33907703e-02 5.55075109e-01 -5.19029647e-02
-1.26508617e+00 1.46461517e-01 -5.78774869e-01 -5.95687807e-01
-6.54841900e-01 -1.11353751e-02 -3.04927528e-01 5.44057339e-02
6.26971126e-01 4.94188160e-01 -2.50545382e-01 -6.76995516e-01
5.20827472e-01 2.58370876e-01 6.24507189e-01 -4.27778721e-01
7.86162078e-01 6.93222880e-01 2.80513279e-02 -8.12014639e-01
-8.33999038e-01 -2.82520771e-01 -8.31788123e-01 -2.93862045e-01
9.89642918e-01 -1.01542974e+00 -4.07530963e-01 6.02426589e-01
-9.08275664e-01 -3.93526435e-01 -5.23510575e-01 3.46133798e-01
-9.58581805e-01 1.89541385e-01 -2.31696144e-01 -6.06938004e-01
-3.07103962e-01 -1.12182188e+00 1.61933494e+00 1.43815428e-01
1.42148539e-01 -7.59360373e-01 -4.29280579e-01 8.23821485e-01
1.09273784e-01 5.79129159e-01 8.92839372e-01 -1.53029382e-01
-1.10524857e+00 -1.28075600e-01 -9.17211771e-02 5.77304363e-01
-3.94277535e-02 -4.60517913e-01 -1.18731868e+00 -1.26660496e-01
5.87385707e-02 -2.32835472e-01 1.11018419e+00 4.50769663e-01
1.18914938e+00 2.24617999e-02 -3.52618665e-01 1.03063977e+00
1.46188641e+00 2.98638046e-01 6.29478872e-01 -1.38324425e-01
9.24653351e-01 3.63954514e-01 4.68907267e-01 1.89256474e-01
3.13913971e-01 7.29557812e-01 9.17470574e-01 -2.66612232e-01
-3.98448199e-01 -5.32274425e-01 3.31494473e-02 3.18792701e-01
-1.42851129e-01 -5.02356470e-01 -7.36716926e-01 4.32986438e-01
-1.31640494e+00 -6.80851221e-01 -1.64342463e-01 1.74607086e+00
3.32730860e-01 1.56329393e-01 -1.10885724e-01 5.41636832e-02
4.36355323e-01 3.73041064e-01 -6.85380578e-01 -2.00612530e-01
-2.80514300e-01 6.50050104e-01 2.91693866e-01 3.77051175e-01
-9.57390130e-01 1.08141696e+00 5.30493116e+00 8.17635357e-01
-9.99166667e-01 -1.98906392e-01 6.74908936e-01 6.35763109e-02
-8.98149848e-01 1.22684706e-02 -7.56521463e-01 6.48748577e-02
4.95183736e-01 3.62212360e-01 1.71699747e-01 9.13526475e-01
-1.26970664e-01 -3.18226218e-01 -1.08982813e+00 1.19379520e+00
3.90850663e-01 -1.32068741e+00 3.43714565e-01 1.72911957e-01
9.06650126e-01 1.67475238e-01 5.07542156e-02 1.71170056e-01
8.76897797e-02 -1.26207674e+00 9.76382434e-01 7.19266891e-01
8.45816731e-01 -8.20811510e-01 5.24515688e-01 6.40272796e-01
-9.51112628e-01 2.07454965e-01 -1.88218206e-01 2.88716435e-01
3.46626073e-01 7.30850101e-01 -9.32391763e-01 1.01904500e+00
9.39603209e-01 7.45077550e-01 -4.63895679e-01 8.70814919e-01
-2.62488127e-01 2.60077685e-01 -6.02679491e-01 2.09400386e-01
3.11352491e-01 -1.53148264e-01 5.98527431e-01 6.74801886e-01
5.34580469e-01 -3.61889973e-02 4.99391891e-02 1.37836611e+00
-2.97963023e-01 -7.63416067e-02 -6.46637022e-01 2.23957717e-01
-1.82890996e-01 1.04686558e+00 -9.23567355e-01 -1.96501076e-01
-4.35030870e-02 1.37006247e+00 -1.57704893e-02 2.76547611e-01
-8.67376328e-01 -2.92713642e-01 3.23857665e-01 3.70707273e-01
7.24460959e-01 -2.76722252e-01 -2.82246202e-01 -1.20123339e+00
1.50528416e-01 -4.30572957e-01 -2.70407349e-02 -1.32443988e+00
-9.93158460e-01 9.42212105e-01 3.29294741e-01 -1.06687701e+00
-3.99826616e-01 -7.46285141e-01 -1.97119579e-01 8.43718469e-01
-1.39825451e+00 -1.37681997e+00 -4.36479717e-01 5.43939412e-01
7.85877168e-01 1.17083810e-01 6.86622083e-01 1.35169506e-01
-9.98780653e-02 2.71928787e-01 -1.32230520e-01 -7.99418762e-02
1.92828700e-01 -1.28211117e+00 8.31643045e-01 6.59926295e-01
5.95616996e-01 1.64278120e-01 4.74740624e-01 -4.90680665e-01
-1.11251938e+00 -1.09024060e+00 5.50099790e-01 -5.47639847e-01
-2.18356505e-01 -6.92940950e-01 -6.45377159e-01 5.33060312e-01
-1.67193934e-01 1.02836438e-01 3.15420657e-01 -2.48464853e-01
-3.80203813e-01 -1.42631471e-01 -1.33150756e+00 4.42827970e-01
1.43967807e+00 -3.82636875e-01 -5.15263259e-01 2.29964077e-01
7.08765924e-01 -5.34396827e-01 -8.96989346e-01 7.78295755e-01
2.87193090e-01 -1.34063482e+00 1.27563202e+00 -3.57148275e-02
4.75887179e-01 -2.97683150e-01 -4.11438346e-01 -1.32158303e+00
-3.16208825e-02 -3.56768847e-01 -7.83118084e-02 8.43038738e-01
1.80181772e-01 -4.26217228e-01 9.28864062e-01 2.34158918e-01
-4.39370066e-01 -7.50876367e-01 -7.73514628e-01 -6.81774974e-01
1.09916657e-01 -8.02947462e-01 7.78071463e-01 4.96185631e-01
-1.01401067e+00 3.57840061e-01 -8.07058811e-02 -3.87699790e-02
6.71470165e-01 6.47694051e-01 7.59411454e-01 -1.35651672e+00
-4.39742297e-01 -3.30691278e-01 -5.05550265e-01 -1.53215432e+00
3.02635252e-01 -1.10671079e+00 1.66442841e-01 -1.72157717e+00
-8.73660371e-02 -4.76279616e-01 1.13227524e-01 4.09045964e-01
1.70285419e-01 6.70846939e-01 1.81820437e-01 -9.84439179e-02
-2.80308634e-01 8.01324964e-01 1.52616417e+00 -3.08299452e-01
-1.51885971e-02 2.24820614e-01 -6.83962464e-01 8.67585063e-01
6.55583024e-01 -1.84985444e-01 -6.26665771e-01 -6.20448291e-01
1.72791407e-01 1.75510481e-01 8.50227952e-01 -1.02063370e+00
-2.93865204e-01 -3.26485634e-02 7.98572421e-01 -1.00038218e+00
8.23979914e-01 -8.35572302e-01 4.10486013e-01 1.03545219e-01
2.16698974e-01 -2.97663808e-01 1.52859688e-01 6.62401259e-01
-1.50872603e-01 -6.42379150e-02 7.18676567e-01 -5.29718876e-01
-1.01671708e+00 5.19078791e-01 4.96403910e-02 -3.87505442e-02
9.60942090e-01 -7.20734179e-01 7.01836050e-02 -2.65231222e-01
-8.76331031e-01 -4.09828275e-02 5.75842857e-01 5.45705497e-01
9.10322785e-01 -1.16477132e+00 -5.85491836e-01 5.82023203e-01
4.74419221e-02 7.76596010e-01 8.21746945e-01 3.54777277e-01
-6.68469429e-01 2.64794201e-01 -1.09321289e-01 -9.59994376e-01
-9.48892713e-01 4.19625610e-01 7.05308914e-01 1.09206632e-01
-9.25845206e-01 1.04222286e+00 7.31306911e-01 -7.32553959e-01
-1.84253991e-01 -4.52398032e-01 2.64803737e-01 -1.95281759e-01
1.79316625e-02 7.49805644e-02 2.56775022e-01 -9.28943396e-01
-2.51180053e-01 7.19954610e-01 1.79906026e-01 -1.46912768e-01
1.55766833e+00 -9.44935437e-03 1.32961959e-01 2.15973854e-01
1.31400490e+00 -1.62026346e-01 -1.60291409e+00 -1.98392153e-01
-5.11586666e-01 -5.16194344e-01 2.80795097e-01 -1.03526008e+00
-1.26088631e+00 1.01680851e+00 4.47872609e-01 -1.99215133e-02
1.24354255e+00 4.24720079e-01 1.11208439e+00 1.90379292e-01
7.39443481e-01 -7.65506923e-01 1.15380406e-01 4.75222260e-01
8.65038216e-01 -9.65531826e-01 -1.60015300e-01 -7.07095981e-01
-2.93545812e-01 9.56902206e-01 6.11365855e-01 -2.45425850e-01
6.06165349e-01 -2.21313499e-02 -9.17064771e-02 -4.34443831e-01
-3.20250779e-01 -2.58930624e-01 6.48042917e-01 7.42715538e-01
6.65107369e-02 1.02890626e-01 4.43901390e-01 1.99100748e-01
-5.27548909e-01 -3.32908362e-01 2.55979687e-01 7.36680806e-01
-2.69195616e-01 -9.30411994e-01 -2.64679581e-01 3.40897411e-01
-3.21210295e-01 -5.33527359e-02 -3.57920289e-01 8.14844668e-01
5.26542485e-01 5.65723956e-01 2.99519151e-01 -3.25030893e-01
5.06726921e-01 1.19182682e-02 7.98420310e-01 -6.45943284e-01
-4.03904825e-01 2.90464550e-01 -1.79736629e-01 -6.43276334e-01
-4.88582700e-01 -4.11653817e-01 -1.17675304e+00 -1.60450768e-02
-2.21802175e-01 -2.39950698e-02 7.46178269e-01 8.78753185e-01
1.82981387e-01 5.70062280e-01 7.03409255e-01 -1.10528255e+00
-1.37998581e-01 -6.66484296e-01 -8.00713241e-01 3.21770012e-01
2.43417487e-01 -6.53148651e-01 -2.60361761e-01 -8.37390199e-02] | [8.560770988464355, -3.2679786682128906] |
f1b83af9-924c-4601-96c2-f7d6357162da | distance-surface-for-event-based-optical-flow | 2003.12680 | null | https://arxiv.org/abs/2003.12680v1 | https://arxiv.org/pdf/2003.12680v1.pdf | Distance Surface for Event-Based Optical Flow | We propose DistSurf-OF, a novel optical flow method for neuromorphic cameras. Neuromorphic cameras (or event detection cameras) are an emerging sensor modality that makes use of dynamic vision sensors (DVS) to report asynchronously the log-intensity changes (called "events") exceeding a predefined threshold at each pixel. In absence of the intensity value at each pixel location, we introduce a notion of "distance surface"---the distance transform computed from the detected events---as a proxy for object texture. The distance surface is then used as an input to the intensity-based optical flow methods to recover the two dimensional pixel motion. Real sensor experiments verify that the proposed DistSurf-OF accurately estimates the angle and speed of each events. | ['Mohammed Almatrafi', 'Raymond Baldwin', 'Kiyoharu Aizawa', 'Keigo Hirakawa'] | 2020-03-28 | null | null | null | null | ['event-based-optical-flow'] | ['computer-vision'] | [ 5.81760406e-01 -5.80328286e-01 1.90844476e-01 -1.95183367e-01
-2.24199910e-02 -8.35409522e-01 4.79398191e-01 -2.54924633e-02
-9.32836294e-01 6.85852110e-01 -1.38638675e-01 3.76475662e-01
1.66915745e-01 -6.60662353e-01 -8.99811208e-01 -8.70158255e-01
-1.10872343e-01 -3.25154930e-01 6.19906843e-01 4.17274773e-01
8.68156850e-01 5.56350589e-01 -1.54649138e+00 1.21341690e-01
1.15419850e-01 1.38669240e+00 2.57691085e-01 1.00747275e+00
8.95683595e-04 5.88315129e-01 -7.18310893e-01 1.13538057e-01
3.34259629e-01 -2.29955062e-01 -1.49911791e-01 -1.55116200e-01
7.00069487e-01 -6.16641223e-01 -7.30261385e-01 1.37134361e+00
1.49668172e-01 -1.90990642e-01 3.73564243e-01 -1.25794554e+00
-2.29032069e-01 -7.81528801e-02 -4.57178503e-01 1.12711465e+00
7.26706624e-01 3.70670408e-01 3.86431009e-01 -8.45165372e-01
1.18379045e+00 9.17195022e-01 5.04214168e-01 6.67739451e-01
-1.20791984e+00 -4.67588484e-01 -1.79692030e-01 3.84643376e-01
-9.74198341e-01 -5.56274712e-01 9.35884058e-01 -5.39295971e-01
8.71162415e-01 -5.21922037e-02 1.06133330e+00 9.22229588e-01
8.39969933e-01 2.53919929e-01 1.13984227e+00 1.14182690e-02
6.83848679e-01 -5.25553107e-01 2.64950544e-01 7.71214426e-01
2.89877713e-01 2.34129876e-01 -1.06354761e+00 -5.86328879e-02
1.46076119e+00 4.45336550e-02 -7.32604861e-01 -1.09652176e-01
-1.38562036e+00 2.34710887e-01 1.95049658e-01 -1.74471125e-01
-3.78872156e-01 7.86422789e-01 1.49546281e-01 1.86333001e-01
-6.32860810e-02 -6.02878928e-02 6.08242601e-02 -7.99716532e-01
-6.65660381e-01 -6.35473952e-02 7.82528698e-01 6.60895705e-01
7.74643362e-01 1.67468581e-02 -1.55390024e-01 -1.60802212e-02
2.86114603e-01 7.20384538e-01 6.75233722e-01 -1.38104010e+00
2.20272779e-01 6.48695529e-01 2.43721992e-01 -9.51960266e-01
-3.53720449e-02 3.40639532e-01 -4.16079998e-01 6.34140730e-01
7.27483809e-01 -2.98251491e-02 -7.66382992e-01 1.66116142e+00
2.67865747e-01 8.14068913e-01 -2.61894286e-01 1.20848501e+00
2.16170728e-01 5.92855275e-01 -4.14003521e-01 -5.93030632e-01
1.19423020e+00 -1.77433550e-01 -6.45773709e-01 -7.90506005e-02
-1.42886534e-01 -3.01516086e-01 6.60981357e-01 5.67431808e-01
-1.20163023e+00 -2.39240795e-01 -1.21427596e+00 3.55306230e-02
-2.96188921e-01 -8.55225548e-02 4.28166509e-01 5.28998077e-01
-1.04943538e+00 5.57159901e-01 -1.20375788e+00 -1.09529473e-01
3.98003042e-01 4.37826604e-01 -2.99526066e-01 3.11123639e-01
-4.74823982e-01 3.27876449e-01 -2.93999493e-01 9.26213115e-02
-1.09406006e+00 -6.19203985e-01 -4.20652837e-01 -3.25699180e-01
-3.00378174e-01 -3.56465399e-01 9.88857567e-01 -9.84299958e-01
-1.71255088e+00 1.05366766e+00 -6.21319592e-01 -6.15423143e-01
4.32242572e-01 -1.27617523e-01 -1.44248366e-01 7.66776502e-01
-3.75085138e-02 5.57816029e-01 1.06825960e+00 -6.71092927e-01
-5.42995036e-01 -6.62616074e-01 -4.34409641e-02 -2.25124359e-02
-2.75436312e-01 -3.61049548e-02 -2.94487327e-01 -1.92065090e-01
2.04581097e-01 -7.82791436e-01 1.30027682e-01 8.85928988e-01
-3.12663704e-01 3.01716793e-02 1.35117114e+00 -6.57374933e-02
9.86657858e-01 -2.12530351e+00 -1.99216977e-02 -1.20314673e-01
2.38894045e-01 1.58512533e-01 2.11348265e-01 -1.11912385e-01
2.45897055e-01 -5.49928129e-01 -2.37613678e-01 -1.47998005e-01
-7.40298152e-01 1.29731596e-01 -5.76773584e-01 1.00715005e+00
1.57737568e-01 5.29957473e-01 -1.17672753e+00 -1.59642249e-01
5.40995657e-01 6.55855834e-01 -1.81216404e-01 1.84565172e-01
-2.49654725e-01 5.51209211e-01 -3.21733475e-01 5.09415925e-01
6.92408025e-01 7.28100166e-03 -2.07626343e-01 -2.90146798e-01
-6.03732467e-01 9.83622000e-02 -1.35909712e+00 1.75165498e+00
2.83405911e-02 1.16888940e+00 -1.50485113e-01 -6.22296512e-01
9.69402432e-01 3.47809196e-02 7.85725892e-01 -7.27190673e-01
3.10749084e-01 5.85740097e-02 -2.86225736e-01 -5.67433119e-01
3.04909617e-01 5.12075484e-01 2.63059616e-01 3.13482732e-01
-7.04829246e-02 1.86557233e-01 1.55313298e-01 5.10402657e-02
1.72667503e+00 1.93935573e-01 -8.43581185e-02 -1.42523184e-01
3.82041603e-01 -2.71959573e-01 8.51282537e-01 5.94181061e-01
-7.51578867e-01 7.00256765e-01 4.69156027e-01 -5.32857656e-01
-9.90661800e-01 -1.48230970e+00 -2.28766397e-01 2.28501201e-01
6.50345683e-01 -2.35355332e-01 -9.64474976e-01 -1.39616027e-01
1.40894607e-01 -1.36653602e-01 -5.33086002e-01 3.70135568e-02
-5.79786718e-01 -5.19532800e-01 4.75145727e-01 5.05476713e-01
5.83013415e-01 -8.63204777e-01 -1.92265344e+00 4.16357368e-01
6.83309184e-03 -1.48112309e+00 -5.61734200e-01 1.71691358e-01
-1.06459069e+00 -1.17400002e+00 -5.05272985e-01 -5.26340067e-01
8.20980966e-01 2.90568352e-01 4.07689333e-01 -5.75933814e-01
-7.27707326e-01 8.27587426e-01 7.61011094e-02 -1.39376089e-01
1.98976979e-01 -1.00191689e+00 4.27223779e-02 5.26247740e-01
3.97025079e-01 -9.46826041e-01 -1.28390133e+00 1.39215291e-01
-9.06011701e-01 -1.34784088e-01 -1.28574520e-01 2.12526083e-01
1.02373314e+00 -7.36313641e-01 6.27041236e-02 -3.59620452e-01
3.97942930e-01 -1.47903144e-01 -9.89185452e-01 -9.66994241e-02
-3.70173991e-01 1.16759926e-01 6.83323205e-01 -1.03352118e+00
-5.37156701e-01 2.95230657e-01 7.28848279e-01 -7.79209912e-01
-1.69794649e-01 -1.19358599e-01 2.31374979e-01 -3.06886315e-01
6.26679957e-01 3.21300119e-01 -2.28935406e-01 -7.08492845e-02
8.93660933e-02 5.43519318e-01 1.14710331e+00 -1.36826903e-01
3.71991657e-02 1.53118694e+00 3.58978957e-01 -8.76808167e-01
-1.16423450e-01 -5.22812486e-01 -5.11736333e-01 -8.00318718e-01
9.11553383e-01 -7.13300824e-01 -1.19331670e+00 1.10887754e+00
-1.61198688e+00 -2.94732451e-01 -2.80706495e-01 4.73163515e-01
-6.87112689e-01 3.26840669e-01 -7.99545705e-01 -9.35541868e-01
-3.51902992e-01 -1.00091803e+00 1.12841308e+00 8.03191483e-01
1.39299929e-01 -8.44761312e-01 3.82197261e-01 -3.69997382e-01
2.86836445e-01 5.70386171e-01 2.60760218e-01 3.07637006e-01
-8.73595417e-01 -3.40188593e-01 -1.44287065e-01 -5.73434494e-03
-1.89092364e-02 3.48823547e-01 -1.18620706e+00 -1.58901945e-01
2.46676907e-01 1.56980321e-01 6.24158859e-01 6.09703302e-01
1.16227889e+00 4.72894311e-02 -4.36732620e-01 8.32308471e-01
1.77073491e+00 4.75680292e-01 8.10229301e-01 2.20568687e-01
4.63272870e-01 -7.77204856e-02 -3.61387432e-02 9.67168272e-01
1.42489433e-01 7.34353364e-01 6.51603341e-01 5.59159696e-01
-3.50514233e-01 -1.37192652e-01 8.76028478e-01 5.42603433e-01
-2.06784829e-01 -7.89773911e-02 -4.14091438e-01 5.07201612e-01
-1.73310125e+00 -9.22781646e-01 -3.76053095e-01 2.74682641e+00
8.29527617e-01 1.01468682e-01 -4.71809395e-02 1.36603668e-01
1.00241756e+00 1.04581073e-01 -1.00860131e+00 -2.77192414e-01
-3.19834232e-01 3.62626702e-01 8.30578744e-01 2.48959661e-01
-5.61904609e-01 5.44823170e-01 6.11130333e+00 -1.74117833e-01
-1.61225307e+00 -6.80348799e-02 -6.30621761e-02 -8.86249170e-02
-1.05752893e-01 2.68637180e-01 -8.47795904e-01 1.06303310e+00
9.38351333e-01 -1.09961018e-01 5.99259555e-01 3.11279267e-01
5.21338046e-01 -8.24996829e-01 -1.05253959e+00 1.37868142e+00
1.11096017e-01 -1.38105214e+00 -1.47071108e-01 9.93762985e-02
3.95293862e-01 -7.10642803e-03 2.57756095e-02 -8.14106166e-01
-9.84376892e-02 -2.76404202e-01 8.98538470e-01 9.70140576e-01
1.04330432e+00 -9.52915326e-02 -1.10509947e-01 -5.63132204e-02
-1.25791407e+00 -1.56390980e-01 -5.18351257e-01 -1.71029925e-01
3.63691479e-01 9.70047176e-01 -5.07608831e-01 -4.66847003e-01
7.78284550e-01 1.01724434e+00 -1.37075663e-01 1.25991333e+00
-3.91080305e-02 5.84691048e-01 -5.68786502e-01 -1.38021648e-01
-1.94299161e-01 -2.61132598e-01 1.03772867e+00 1.00259209e+00
4.56011742e-01 -3.90518643e-02 -4.50244308e-01 1.17009640e+00
7.13039283e-03 -6.48546636e-01 -6.57994866e-01 5.69627881e-02
7.73276031e-01 1.19788349e+00 -7.52122760e-01 -1.42409056e-01
-4.55321431e-01 1.51523268e+00 1.15443937e-01 3.29811722e-01
-5.88725448e-01 -6.44299388e-01 8.83311749e-01 1.18378773e-01
1.73075214e-01 -5.70323646e-01 -1.70687482e-01 -1.19165790e+00
3.15093815e-01 1.57767668e-01 1.18686065e-01 -1.13819087e+00
-6.59737527e-01 4.61376868e-02 -5.31476021e-01 -1.41912687e+00
-6.18011691e-02 -7.99026072e-01 -7.98628926e-01 3.23226810e-01
-1.48807037e+00 -3.68377835e-01 -7.57787347e-01 1.15316260e+00
2.30424166e-01 1.73356563e-01 5.42286217e-01 -3.26155387e-02
-3.56119871e-01 9.45221856e-02 7.33664706e-02 2.61217296e-01
5.68343222e-01 -1.06514132e+00 2.00183690e-01 1.13552272e+00
1.54576212e-01 2.36771822e-01 7.13929355e-01 -6.68878436e-01
-2.17144823e+00 -6.94588184e-01 3.54196906e-01 -3.53902072e-01
8.10221314e-01 -3.30737621e-01 -5.58285236e-01 5.20496190e-01
-1.02139622e-01 6.42358363e-01 1.02095425e-01 -1.13039494e+00
-5.05014837e-01 -6.06214941e-01 -1.35146797e+00 3.47028941e-01
1.17833006e+00 -7.87508726e-01 -3.51378828e-01 -5.31181432e-02
3.57371747e-01 -3.52066725e-01 -6.95636988e-01 -2.07515135e-02
8.02649319e-01 -1.21776092e+00 6.87861919e-01 -1.45209096e-02
1.71589404e-01 -5.00548899e-01 -2.76587866e-02 -8.07649374e-01
2.97765076e-01 -1.04165864e+00 -4.23530489e-01 7.48484075e-01
-2.16680631e-01 -7.06500113e-01 9.28303659e-01 4.13455039e-01
1.30924717e-01 -1.94887266e-01 -1.56321406e+00 -6.96570098e-01
-6.06500924e-01 -5.92823446e-01 -1.43934444e-01 4.68499511e-01
2.58789241e-01 -3.44426900e-01 1.90733418e-01 3.07444543e-01
1.08493590e+00 -1.25528976e-01 2.95384765e-01 -1.04848731e+00
-1.03663847e-01 -2.22383395e-01 -1.38040781e+00 -1.24328244e+00
-2.21031383e-01 -4.49452221e-01 1.61240831e-01 -1.00578272e+00
7.05747232e-02 2.66901940e-01 -3.31978232e-01 -3.63571383e-02
1.80841431e-01 1.82919800e-01 -5.33573329e-02 3.53255093e-01
-4.20956939e-01 2.26802565e-02 9.59891856e-01 2.81971365e-01
-3.49753797e-01 -3.21043551e-01 2.94145375e-01 6.18123651e-01
3.02314639e-01 -3.37696731e-01 -2.30568960e-01 -5.36104143e-01
6.52864277e-02 3.38494867e-01 8.88881087e-01 -1.54419839e+00
8.52050006e-01 2.34910641e-02 5.02303183e-01 -1.99697003e-01
4.63800192e-01 -6.04078650e-01 6.69108517e-03 4.44137335e-01
-2.82882631e-01 5.49895823e-01 -7.26577193e-02 9.49605823e-01
1.06820352e-01 -5.89607060e-02 7.87850082e-01 -5.13872243e-02
-9.06452298e-01 4.21904236e-01 -6.04733407e-01 1.11247562e-01
1.11605752e+00 -7.21038818e-01 -8.31663251e-01 -1.57516554e-01
-2.79209167e-01 -4.11492676e-01 6.17967546e-01 1.19377993e-01
1.15960932e+00 -1.14182556e+00 -6.38625771e-02 4.73984003e-01
1.42292334e-02 -1.84013128e-01 -1.14423983e-01 8.58322144e-01
-6.73003674e-01 6.33514598e-02 -5.57810605e-01 -9.20987964e-01
-9.88623619e-01 1.38053134e-01 5.68064094e-01 4.64620918e-01
-8.78369272e-01 8.32795680e-01 -2.20915422e-01 7.43260264e-01
3.08609486e-01 -4.90775973e-01 1.47134215e-01 -2.99686521e-01
9.13305044e-01 5.82550645e-01 -1.53314590e-01 -4.20282304e-01
-4.36205357e-01 9.64473069e-01 4.03320551e-01 -8.48163009e-01
1.05977201e+00 -2.59518296e-01 -1.24029212e-01 8.27730834e-01
1.22559202e+00 -2.96395659e-01 -2.09341311e+00 1.91615727e-02
-2.33879909e-02 -5.77125132e-01 1.75829813e-01 -4.13037181e-01
-1.12828469e+00 1.10139525e+00 1.14874172e+00 1.01431794e-01
1.09289420e+00 -2.66443521e-01 1.01036561e+00 1.09895147e-01
5.78042090e-01 -1.23912537e+00 1.79055706e-01 2.31856570e-01
2.46525198e-01 -4.61487889e-01 -5.60086370e-01 -1.91879272e-01
6.01501539e-02 1.64232326e+00 4.89664406e-01 -6.28865361e-01
5.68397343e-01 8.57695162e-01 -3.47248167e-02 -2.61099860e-02
-7.87927568e-01 3.62598374e-02 -4.76715565e-01 7.88099468e-01
-1.20097280e-01 -1.50015846e-01 -1.55189663e-01 -1.47446185e-01
3.25930059e-01 5.94715774e-01 9.47897315e-01 1.10381830e+00
-6.85285151e-01 -4.75712568e-01 -3.83330554e-01 2.51331806e-01
-2.43161246e-01 3.08318347e-01 -2.57720947e-01 -6.15083566e-03
8.49309117e-02 7.80599892e-01 7.06116796e-01 -4.03491706e-01
3.09740752e-01 -1.27637595e-01 9.67336416e-01 -7.22060576e-02
-5.23323417e-02 -1.70297518e-01 -5.47671497e-01 -1.12864006e+00
-6.59650266e-01 -7.13294566e-01 -1.69439995e+00 3.75460042e-03
1.46856785e-01 -4.56116557e-01 1.05972588e+00 7.79945731e-01
4.13961321e-01 -6.06979169e-02 8.54722857e-01 -6.52541935e-01
-1.46834105e-01 -3.46400529e-01 -6.18100464e-01 7.24764109e-01
6.71912670e-01 -4.98681277e-01 -8.38867009e-01 5.14312863e-01] | [8.666570663452148, -1.2707784175872803] |
7b09db9b-4236-4ee2-9bf1-aca1998d645e | towards-classification-of-legal | null | null | https://aclanthology.org/2022.csrnlp-1.8 | https://aclanthology.org/2022.csrnlp-1.8.pdf | Towards Classification of Legal Pharmaceutical Text using GAN-BERT | Pharmaceutical text classification is an important area of research for commercial and research institutions working in the pharmaceutical domain. Addressing this task is challenging due to the need of expert verified labelled data which can be expensive and time consuming to obtain. Towards this end, we leverage predictive coding methods for the task as they have been shown to generalise well for sentence classification. Specifically, we utilise GAN-BERT architecture to classify pharmaceutical texts. To capture the domain specificity, we propose to utilise the BioBERT model as our BERT model in the GAN-BERT framework. We conduct extensive evaluation to show the efficacy of our approach over baselines on multiple metrics. | ['John P. McCrae', 'Vall Herard', 'John Mariano', 'Jay Megaro', 'Michaela Comerford', 'Arindam Paul', 'Atul Kr. Ojha', 'Bernardo Stearns', 'Rajdeep Sarkar', 'Tapan Auti'] | null | null | null | null | csrnlp-lrec-2022-6 | ['sentence-classification'] | ['natural-language-processing'] | [ 5.17977595e-01 2.34439984e-01 -8.90914351e-02 -4.49999005e-01
-1.03675568e+00 -5.83517134e-01 7.09919870e-01 2.81798810e-01
-2.70682693e-01 8.32329631e-01 2.71510422e-01 -6.82741225e-01
1.92123074e-02 -4.01207298e-01 -7.44354784e-01 -4.64048237e-01
2.17783391e-01 5.47839105e-01 -2.53822744e-01 -1.17672838e-01
1.64382249e-01 3.09198707e-01 -9.26144540e-01 7.67423332e-01
9.25396085e-01 1.02873504e+00 -8.77020285e-02 6.54121518e-01
9.59999785e-02 1.05866075e+00 -5.62151909e-01 -7.42350280e-01
1.91949040e-01 -5.84730744e-01 -9.99047935e-01 -1.96616221e-02
4.33991291e-02 -3.65975127e-02 2.54183382e-01 8.69238555e-01
4.08635885e-01 -2.65763793e-02 1.04620886e+00 -7.84170210e-01
-4.17483777e-01 5.22643387e-01 -2.66995519e-01 1.91201627e-01
3.69262636e-01 1.80525437e-01 1.20553672e+00 -6.38813496e-01
6.33000672e-01 1.00855231e+00 5.50165534e-01 5.14524937e-01
-1.44002056e+00 -5.05500853e-01 5.07915616e-02 -1.81994971e-03
-9.70378458e-01 -5.74020505e-01 6.33252025e-01 -6.31898403e-01
9.81697738e-01 1.03529729e-01 3.72449487e-01 1.50701678e+00
4.70491409e-01 1.00148523e+00 1.28042173e+00 -2.88602084e-01
4.19129968e-01 1.96272358e-01 1.51030067e-02 2.84432948e-01
4.53380160e-02 -3.72031368e-02 -2.91212052e-01 2.75718700e-02
3.04841965e-01 4.01527733e-02 -3.48019689e-01 -2.80024439e-01
-7.04653978e-01 1.32693827e+00 4.83555287e-01 2.15400666e-01
-6.22155368e-01 3.37013341e-02 6.66244566e-01 4.87087607e-01
8.56879830e-01 8.36153209e-01 -5.64627230e-01 -4.15238976e-01
-1.06650531e+00 1.50719747e-01 9.25809145e-01 8.51184666e-01
-2.58242544e-02 -3.12430441e-01 -2.37095773e-01 7.36849844e-01
2.49447256e-01 -1.87489301e-01 4.97717679e-01 -3.75918239e-01
5.58541477e-01 6.62330449e-01 -1.68856770e-01 -5.41683018e-01
-3.68274897e-01 -5.80457449e-01 -9.50925469e-01 -1.36598840e-01
1.16668157e-01 -8.87150839e-02 -1.10937405e+00 1.24564731e+00
6.43560663e-02 -3.69472951e-02 3.50633025e-01 5.58535337e-01
7.42786169e-01 5.29067099e-01 4.43495035e-01 -1.29402429e-01
1.31668150e+00 -1.07705414e+00 -4.82149452e-01 -1.63425088e-01
8.43065262e-01 -6.24845982e-01 9.40332294e-01 5.57357907e-01
-8.85275304e-01 -1.01327486e-01 -9.24255967e-01 -1.79685771e-01
-3.71747434e-01 1.36974603e-01 5.42377889e-01 7.39521801e-01
-6.21131480e-01 6.43909574e-01 -9.83991385e-01 -2.14693263e-01
9.75762427e-01 2.78467298e-01 -1.75841138e-01 -2.73246229e-01
-1.10217464e+00 9.85499799e-01 5.05016446e-01 2.45244041e-01
-7.15752840e-01 -7.08533466e-01 -7.60795295e-01 1.09679051e-01
1.81805924e-01 -7.83692718e-01 1.46620226e+00 -7.85979867e-01
-1.54762638e+00 6.31180763e-01 1.33205950e-01 -7.79316723e-01
7.63030469e-01 -7.02460781e-02 -7.23790303e-02 1.01968542e-01
-1.46616802e-01 4.70675111e-01 4.98737425e-01 -8.38911653e-01
-4.41857338e-01 -1.72868893e-01 1.90746844e-01 5.53222299e-02
-1.94981068e-01 3.13263573e-02 -3.37970220e-02 -7.27195501e-01
-4.08901513e-01 -9.59435165e-01 -4.20328468e-01 -5.23297727e-01
-5.07994533e-01 -2.52476454e-01 4.52418119e-01 -9.32689786e-01
1.07079458e+00 -2.03852177e+00 1.57688707e-01 1.07954368e-01
2.75131226e-01 4.46258426e-01 1.13595814e-01 5.07560134e-01
-2.20880862e-02 1.15968466e-01 -3.17516476e-01 -2.41574138e-01
-2.64778882e-02 2.43458390e-01 -2.27013066e-01 2.77267516e-01
5.65260470e-01 1.12864935e+00 -7.08390653e-01 -2.07060456e-01
1.60140842e-02 5.92841804e-01 -7.05908477e-01 2.45940492e-01
-5.24607837e-01 6.54249847e-01 -5.32953084e-01 5.68539202e-01
1.85635671e-01 -5.79114258e-01 3.30891609e-01 -1.05926551e-01
2.28015572e-01 7.94477642e-01 -4.60532963e-01 1.58005059e+00
-5.66167057e-01 2.99290448e-01 -1.32190704e-01 -1.29171598e+00
5.82958162e-01 5.00586033e-01 3.87482971e-01 -5.04884362e-01
3.08103889e-01 3.13888043e-01 1.76169738e-01 -5.10426998e-01
1.59886390e-01 -6.62895977e-01 1.72045622e-02 1.92673698e-01
6.45020083e-02 -1.04701206e-01 1.52225330e-01 9.70313773e-02
1.24260509e+00 2.36064717e-01 4.98412132e-01 -1.31321043e-01
4.46512073e-01 1.02404192e-01 1.92261890e-01 4.11287338e-01
1.71743624e-03 3.77903759e-01 6.76028490e-01 -3.03670675e-01
-8.34031463e-01 -3.80654156e-01 -2.85598189e-01 9.56740141e-01
-5.26746094e-01 -4.88412738e-01 -6.94734454e-01 -1.12136221e+00
-1.26762807e-01 5.98714888e-01 -6.43815160e-01 -3.71705890e-02
-3.26811016e-01 -8.52562964e-01 3.50817740e-01 6.49316370e-01
-7.73114190e-02 -9.03285921e-01 -7.46715844e-01 4.95697051e-01
-4.51910496e-02 -1.16059387e+00 -2.09387705e-01 7.90284574e-01
-6.01207852e-01 -1.01262689e+00 -9.26442206e-01 -7.49148786e-01
3.05813402e-01 -2.59047508e-01 1.23073864e+00 -1.91860110e-01
-1.97383285e-01 1.90616444e-01 -6.70713186e-01 -8.05973530e-01
-7.50223935e-01 5.11954486e-01 -5.28096139e-01 -1.57192796e-01
3.83591354e-01 -1.88349470e-01 -7.80460298e-01 8.25886354e-02
-1.11861193e+00 1.49569750e-01 5.94187915e-01 1.16215491e+00
3.40400636e-01 -2.06007273e-03 6.75295115e-01 -1.49571931e+00
9.73471224e-01 -6.07633054e-01 -4.60826486e-01 1.29540294e-01
-7.66708970e-01 1.27443880e-01 8.11944067e-01 -1.42812565e-01
-4.49987262e-01 2.27437973e-01 -4.56960589e-01 -1.26833558e-01
-6.21112697e-02 1.15520620e+00 -1.00312456e-01 7.23295733e-02
5.18543303e-01 -4.63120416e-02 1.85693558e-02 -4.64262545e-01
3.04900229e-01 1.05957031e+00 7.14780390e-02 -2.59333938e-01
3.27216715e-01 7.17903972e-02 -2.55895033e-03 -4.09426332e-01
-1.06331301e+00 -5.78031480e-01 -3.64627093e-01 1.75853789e-01
8.86354089e-01 -9.79598224e-01 -5.80996335e-01 2.98421588e-02
-1.14466631e+00 -2.84142047e-01 -3.77322622e-02 3.30365479e-01
-5.67838371e-01 2.87017465e-01 -7.47549772e-01 -7.28862703e-01
-6.20932102e-01 -1.44243157e+00 1.27412832e+00 -3.51787150e-01
-5.17408729e-01 -1.21533585e+00 1.17228217e-02 4.96570915e-01
4.46148008e-01 4.99956518e-01 1.10588932e+00 -1.04776561e+00
-1.83536619e-01 -3.12768281e-01 -1.13725655e-01 4.30644542e-01
-2.62589641e-02 -3.61147106e-01 -1.13986516e+00 -3.87722403e-01
2.78454542e-01 -6.00909173e-01 9.99986649e-01 2.87894309e-01
1.18181479e+00 -4.22033906e-01 -2.84375340e-01 2.44110316e-01
1.24076045e+00 2.00547829e-01 5.46321571e-01 2.77971774e-01
6.23232365e-01 7.77732790e-01 4.33768928e-01 2.65829951e-01
3.82107288e-01 7.72598505e-01 1.30319491e-01 -2.79872298e-01
2.03907505e-01 -5.27418405e-02 2.00255051e-01 8.58939707e-01
1.74175024e-01 -5.04738986e-01 -1.07224131e+00 3.58283728e-01
-1.78618467e+00 -7.89588332e-01 -8.20636600e-02 1.79715621e+00
1.10146105e+00 2.64685929e-01 2.60561794e-01 2.92548299e-01
1.39479354e-01 -2.61107266e-01 -4.66108441e-01 -6.60566747e-01
2.67973721e-01 5.85352123e-01 4.48885530e-01 1.33308589e-01
-1.23967838e+00 7.78375685e-01 6.24445009e+00 7.55231619e-01
-1.30451977e+00 4.74225618e-02 9.28927898e-01 -7.95138851e-02
6.59572612e-03 -3.12271953e-01 -6.72581255e-01 7.53504336e-01
1.45989037e+00 4.59031239e-02 2.98787624e-01 6.78559899e-01
1.59182325e-01 1.44155681e-01 -1.60776043e+00 8.45946670e-01
-3.16020963e-03 -1.20812058e+00 2.43505035e-02 2.48498365e-01
6.09092474e-01 2.66164821e-02 9.88998488e-02 4.25651163e-01
5.77234805e-01 -1.50034797e+00 4.25898731e-01 2.00324923e-01
6.49566531e-01 -5.44712543e-01 9.25910711e-01 3.58764291e-01
-7.24806428e-01 -4.81704362e-02 -2.31472433e-01 5.91924079e-02
7.61115104e-02 5.06569445e-01 -1.29127645e+00 6.89651668e-01
3.79485130e-01 8.76257479e-01 -3.79794449e-01 9.97065365e-01
-1.59840316e-01 7.91571736e-01 -1.09978653e-01 2.31248653e-03
4.78925079e-01 -2.02563882e-01 -6.24631159e-02 1.32988226e+00
1.52574494e-01 -1.58039957e-01 3.13770354e-01 5.91090024e-01
-3.94379348e-01 3.01115513e-01 -5.65357625e-01 -4.95474607e-01
-1.23964787e-01 9.39502120e-01 -6.81497872e-01 -2.51200169e-01
-3.97898585e-01 1.02894962e+00 2.88634956e-01 7.26860017e-02
-6.25923097e-01 -2.41582677e-01 2.41048813e-01 -2.83883512e-02
6.08529449e-01 1.03029311e-01 -2.18553394e-01 -9.74742532e-01
5.73960394e-02 -1.30562925e+00 5.62464774e-01 -4.32954609e-01
-1.48092794e+00 7.09245563e-01 -2.60927528e-01 -1.24308968e+00
-4.10032928e-01 -8.59449565e-01 -1.13458954e-01 7.99012482e-01
-1.85128403e+00 -1.35121250e+00 4.55684662e-02 2.25173503e-01
6.96726143e-01 -8.15879181e-02 8.76111865e-01 3.67452741e-01
-4.65076327e-01 6.20282948e-01 2.76764929e-01 8.37947652e-02
5.29324174e-01 -1.29438543e+00 3.55021447e-01 3.91365916e-01
1.98187768e-01 6.52568519e-01 6.42294049e-01 -4.61982220e-01
-1.26613593e+00 -1.34740269e+00 8.88410211e-01 -6.32242322e-01
7.95397103e-01 -5.74749351e-01 -7.49925971e-01 6.04598999e-01
1.92911014e-01 -3.88050824e-01 1.19047868e+00 8.47291574e-02
-3.11133116e-01 3.17418724e-01 -9.32808995e-01 3.08173537e-01
5.77905595e-01 -5.83488882e-01 -3.97073895e-01 7.00102031e-01
6.22159123e-01 -4.79203522e-01 -1.06675839e+00 2.37409636e-01
3.94345015e-01 -4.90557015e-01 6.80388629e-01 -8.32319498e-01
9.62097406e-01 7.79953152e-02 3.89795825e-02 -1.54643261e+00
-8.66233781e-02 -3.32875431e-01 1.85924292e-01 9.98463929e-01
8.54347348e-01 -6.02817535e-01 8.26141477e-01 6.51793838e-01
-6.57219738e-02 -1.10321105e+00 -9.88677979e-01 -8.12522113e-01
3.64357710e-01 -2.82286376e-01 3.95528674e-01 9.34736371e-01
1.78583235e-01 9.08536434e-01 -3.86168778e-01 -3.65007669e-01
8.08906853e-02 9.37118903e-02 6.18223906e-01 -1.26348364e+00
-5.89748085e-01 -5.57805836e-01 -5.91648996e-01 -9.63934183e-01
2.76399404e-01 -1.15972078e+00 8.57541338e-02 -1.62416220e+00
2.73423791e-01 -2.81123579e-01 -3.38513702e-01 5.31633496e-01
-2.96675444e-01 3.42499197e-01 1.18843488e-01 2.41612047e-01
-6.44600809e-01 5.92693806e-01 9.61446106e-01 -4.50363517e-01
-5.27601503e-02 -2.12237649e-02 -1.03817976e+00 2.19657630e-01
7.80569613e-01 -4.82722074e-01 -4.51296657e-01 -2.14676648e-01
3.66513401e-01 -4.83180210e-02 5.12496233e-02 -7.10006297e-01
-1.23523533e-01 1.38987929e-01 1.61107227e-01 -2.05869541e-01
2.59130925e-01 -9.94442463e-01 2.44812816e-02 4.48231667e-01
-7.03447342e-01 -4.05495726e-02 9.78941917e-02 7.41564155e-01
-4.02254373e-01 -2.94229090e-01 6.06710136e-01 -6.70198724e-02
5.99298701e-02 1.55456111e-01 -3.30503196e-01 4.95614968e-02
1.03211582e+00 -6.37581125e-02 -5.85152917e-02 -3.17987025e-01
-4.65213418e-01 2.88989931e-01 3.19436908e-01 2.06607372e-01
3.65983784e-01 -9.71375108e-01 -6.86502278e-01 -5.92979379e-02
3.59193891e-01 -3.02736331e-02 -2.37465411e-01 1.05346298e+00
-4.92403448e-01 8.20167661e-01 1.79933280e-01 -5.31538367e-01
-1.22684157e+00 6.82575822e-01 2.36816257e-01 -6.65295422e-01
-5.43229699e-01 7.22460985e-01 2.56495625e-01 -3.33558798e-01
1.86931223e-01 -7.17525125e-01 -5.02870679e-01 6.66251332e-02
2.34277233e-01 -1.41805718e-02 5.66624403e-01 -3.10145706e-01
-2.16875091e-01 1.13573737e-01 -4.57361877e-01 1.56862229e-01
1.61472344e+00 2.74518013e-01 1.71222836e-01 1.98361039e-01
1.40569425e+00 -3.13401788e-01 -9.44927037e-01 4.86624055e-02
4.33845490e-01 -1.42573565e-03 1.79088265e-01 -1.09432840e+00
-7.67762780e-01 9.75572348e-01 4.05224770e-01 3.25382799e-01
9.69408691e-01 -1.30535841e-01 7.69162238e-01 7.17877075e-02
5.89296706e-02 -8.16755652e-01 -3.11581403e-01 4.24451590e-01
8.24035645e-01 -1.39187050e+00 -4.06491384e-02 -5.29985547e-01
-7.79228151e-01 8.91456187e-01 -7.42034316e-02 1.52740330e-01
5.51713824e-01 8.95321146e-02 1.17895216e-01 -4.41847742e-01
-8.84292126e-01 -4.84305294e-03 3.36581886e-01 4.03202742e-01
9.88511682e-01 1.26638025e-01 -4.90804672e-01 4.53792006e-01
-2.18994007e-01 2.88577795e-01 3.90057117e-01 1.20549989e+00
1.17858127e-01 -1.39079189e+00 7.22259656e-02 7.54328966e-01
-9.83800352e-01 -3.25071454e-01 -6.77321315e-01 2.85786241e-01
-3.85152727e-01 1.13637054e+00 -3.85184348e-01 -1.43428773e-01
3.28944534e-01 3.17371339e-01 3.35466146e-01 -8.85532796e-01
-9.10267889e-01 3.00170243e-01 2.38603279e-01 -1.49997652e-01
-5.44230342e-01 -5.12061536e-01 -8.61112177e-01 6.44086450e-02
-2.27197289e-01 3.48538727e-01 7.30395794e-01 8.91529739e-01
5.46775162e-01 7.31964707e-01 4.65318918e-01 -3.64432007e-01
-9.06584442e-01 -1.09143233e+00 -1.77115783e-01 5.94457865e-01
2.54289448e-01 -4.89921451e-01 -3.61562110e-02 2.62201339e-01] | [8.605764389038086, 8.498472213745117] |
379dc4b1-de9c-41b9-89b5-aebe178934d9 | multi-object-tracking-and-segmentation-with-a | 2110.11284 | null | https://arxiv.org/abs/2110.11284v2 | https://arxiv.org/pdf/2110.11284v2.pdf | Multi-Object Tracking and Segmentation with a Space-Time Memory Network | We propose a method for multi-object tracking and segmentation based on a novel memory-based mechanism to associate tracklets. The proposed tracker, MeNToS, addresses particularly the long-term data association problem, when objects are not observable for long time intervals. Indeed, the recently introduced HOTA metric (High Order Tracking Accuracy), which has a better alignment than the formerly established MOTA (Multiple Object Tracking Accuracy) with the human visual assessment of tracking, has shown that improvements are still needed for data association, despite the recent improvement in object detection. In MeNToS, after creating tracklets using instance segmentation and optical flow, the proposed method relies on a space-time memory network originally developed for one-shot video object segmentation to improve the association of sequence of detections (tracklets) with temporal gaps. We evaluate our tracker on KITTIMOTS and MOTSChallenge and we show the benefit of our data association strategy with the HOTA metric. Additional ablation studies demonstrate that our approach using a space-time memory network gives better and more robust long-term association than those based on a re-identification network. Our project page is at \url{www.mehdimiah.com/mentos+}. | ['Nicolas Saunier', 'Guillaume-Alexandre Bilodeau', 'Mehdi Miah'] | 2021-10-21 | null | null | null | null | ['multi-object-tracking-and-segmentation'] | ['computer-vision'] | [-1.37099713e-01 -3.36217046e-01 -1.03219196e-01 7.10674152e-02
-3.55815321e-01 -7.26593733e-01 4.64004338e-01 2.70851791e-01
-6.13334119e-01 6.91980600e-01 -5.67324758e-01 -1.23428879e-02
-5.47423005e-01 -4.76286352e-01 -8.07276607e-01 -4.88467187e-01
-3.15421253e-01 7.60052681e-01 1.03484476e+00 1.52325973e-01
1.23440564e-01 7.75062144e-01 -1.90148544e+00 -5.22408523e-02
4.77046400e-01 9.32724893e-01 2.46325910e-01 9.97306883e-01
1.02425501e-01 4.35820788e-01 -6.70772195e-01 -6.09590486e-02
5.27613103e-01 -1.92106694e-01 -5.74990153e-01 -5.43684140e-02
1.16259933e+00 -1.99870542e-01 -2.22040787e-01 9.50327575e-01
4.01358724e-01 6.27928302e-02 2.54533470e-01 -1.59731042e+00
-1.48822755e-01 4.14977610e-01 -5.62396944e-01 7.48830557e-01
1.50229663e-01 2.03020573e-01 5.95724821e-01 -5.52870154e-01
7.07336783e-01 1.09519649e+00 1.25453818e+00 5.00544012e-01
-1.07710683e+00 -8.49669814e-01 6.69040978e-02 4.48825508e-01
-1.44050097e+00 -1.91743225e-01 1.17438197e-01 -6.86597228e-01
5.94371498e-01 3.71815026e-01 8.10748994e-01 7.13625848e-01
1.94957435e-01 5.49020290e-01 9.27451551e-01 -2.53045559e-01
-1.40113279e-01 2.78558791e-01 5.01662314e-01 6.92035973e-01
7.77809978e-01 5.69637656e-01 -5.32874405e-01 -6.78508431e-02
9.43311453e-01 -4.97604981e-02 1.23181373e-01 -5.10939717e-01
-1.46804893e+00 4.43626016e-01 8.98861736e-02 6.28277838e-01
-2.15643123e-01 4.01046872e-01 2.42577985e-01 2.27446228e-01
1.07299857e-01 2.37466559e-01 -2.36146167e-01 -1.82921335e-01
-1.29464877e+00 3.01422358e-01 5.60265183e-01 1.08856380e+00
6.96925998e-01 1.08045191e-01 -4.74036455e-01 1.68468118e-01
4.20252800e-01 7.28063703e-01 3.15376252e-01 -8.67274404e-01
3.95349041e-03 5.56491971e-01 3.36457223e-01 -9.74742234e-01
-7.95893669e-01 -6.01885200e-01 -3.19480777e-01 8.32832158e-01
7.92826951e-01 -1.65754020e-01 -8.60384047e-01 1.64804411e+00
6.56193256e-01 6.50813103e-01 -2.16795921e-01 8.70312035e-01
6.02014303e-01 1.79664850e-01 8.37706402e-02 -2.73611665e-01
1.28628910e+00 -8.05866897e-01 -8.68881941e-01 1.55283093e-01
6.33989751e-01 -8.62003505e-01 6.37886301e-02 2.78806567e-01
-1.01253915e+00 -1.09925163e+00 -1.08765936e+00 5.84625483e-01
-5.09014547e-01 1.08392313e-01 4.79159355e-01 9.68798935e-01
-1.13045585e+00 9.37875569e-01 -9.59509134e-01 -8.91379714e-01
3.14164877e-01 7.11448729e-01 -1.88159317e-01 6.37876689e-01
-8.63264143e-01 1.02968633e+00 6.64729953e-01 9.35161188e-02
-9.71274734e-01 -5.76557100e-01 -4.29454863e-01 -2.95430481e-01
5.48240185e-01 -5.66257954e-01 8.88298094e-01 -4.94081855e-01
-1.19705021e+00 7.23629653e-01 -6.27973466e-04 -8.92419755e-01
7.28198409e-01 -4.11145508e-01 -6.81896031e-01 1.23155385e-01
1.14677370e-01 7.73503304e-01 9.91741359e-01 -1.03887248e+00
-1.06013846e+00 -1.87351331e-01 -1.05248384e-01 -1.69404641e-01
-2.23382324e-01 1.01316951e-01 -2.33134598e-01 -5.31534791e-01
-4.19134088e-02 -1.12103319e+00 -1.45098895e-01 2.52653092e-01
2.13425066e-02 -2.39736900e-01 1.39243698e+00 -3.82429630e-01
1.27193463e+00 -1.98214197e+00 -1.72116175e-01 3.05001616e-01
4.10245389e-01 8.03021908e-01 -4.72439565e-02 9.27770957e-02
-6.17963299e-02 -2.86424458e-01 1.80776536e-01 -2.38217428e-01
-1.56392738e-01 8.83543193e-02 -6.47755107e-03 6.89898849e-01
-2.77195394e-01 7.39633381e-01 -9.00451064e-01 -8.25797677e-01
5.96891046e-01 2.44050786e-01 -1.24457426e-01 8.47448688e-03
-1.09511182e-01 6.04424298e-01 -2.40217030e-01 5.49002349e-01
7.33903885e-01 -2.72803038e-01 -1.82601392e-01 -3.06538969e-01
-6.98808253e-01 -5.72735608e-01 -1.69046867e+00 1.55525959e+00
2.06679747e-01 7.87663698e-01 -7.85026848e-02 -5.48200250e-01
9.61250424e-01 3.96119565e-01 1.05150235e+00 -4.78764296e-01
1.59017205e-01 1.96829811e-01 9.30247828e-02 -4.35082674e-01
7.98468292e-01 3.00532699e-01 3.40145350e-01 3.65931988e-01
2.11887047e-01 4.89652574e-01 6.49157047e-01 2.99199581e-01
1.15566254e+00 5.19885421e-01 8.22503716e-02 -3.73336405e-01
6.01916075e-01 2.98867345e-01 6.50806606e-01 1.17725003e+00
-6.56684935e-01 4.95010346e-01 -2.54324675e-01 -3.67828459e-01
-9.82342482e-01 -9.78047371e-01 -2.77670979e-01 9.37008500e-01
5.79191625e-01 -5.02870798e-01 -6.15275204e-01 -5.88907957e-01
2.37031117e-01 2.21528057e-02 -6.72425032e-01 1.12060979e-01
-8.29917133e-01 -4.58439231e-01 7.44293213e-01 6.47228360e-01
3.40952188e-01 -9.19009864e-01 -1.06037700e+00 4.13621157e-01
1.50705174e-01 -1.12355101e+00 -5.67181826e-01 -5.44332489e-02
-9.74822879e-01 -1.27063680e+00 -6.65669024e-01 -3.83730978e-01
3.38105232e-01 2.10408673e-01 8.43324423e-01 3.88116628e-01
-5.45476973e-01 8.78063917e-01 -3.57033104e-01 -3.90116692e-01
-3.82360607e-01 -1.91474676e-01 4.34657395e-01 2.42918003e-02
2.55311072e-01 -2.99456626e-01 -5.02208650e-01 7.71941900e-01
-6.27512634e-01 -1.92125484e-01 4.64015663e-01 3.47886533e-01
3.66844118e-01 -2.60996372e-01 4.51955706e-01 -5.16809106e-01
-1.77009612e-01 -1.83432698e-01 -1.16292346e+00 3.07910413e-01
-7.13036716e-01 -8.70660245e-02 -8.65704566e-02 -7.85277307e-01
-6.23863816e-01 3.25614244e-01 1.90236136e-01 -7.96373606e-01
-1.74802199e-01 -2.26285696e-01 2.46524230e-01 -6.48998260e-01
6.94425762e-01 -1.09464303e-01 1.71951219e-01 -5.30921757e-01
3.44198644e-01 2.07232133e-01 6.43545985e-01 -3.41585100e-01
1.01727462e+00 7.98554838e-01 3.23806256e-01 -5.70114791e-01
-5.64770818e-01 -8.38324487e-01 -1.05400884e+00 -8.22372139e-01
1.00208211e+00 -6.65212512e-01 -1.18505573e+00 4.83201414e-01
-1.23701715e+00 1.48536935e-02 -6.41203582e-01 9.02046800e-01
-5.95036864e-01 4.67017382e-01 -3.22865188e-01 -1.02628696e+00
-1.83066547e-01 -7.20533013e-01 8.27838242e-01 5.10736346e-01
-1.79866493e-01 -1.05603170e+00 3.72337848e-01 1.18674055e-01
4.78690058e-01 3.94206971e-01 -9.29137841e-02 -6.48621142e-01
-1.07877135e+00 -9.90526006e-02 -2.01825038e-01 -2.83476729e-02
-9.51678231e-02 8.61046091e-02 -8.31366897e-01 -6.13075972e-01
-1.63809881e-01 3.73197585e-01 7.41739690e-01 5.74743152e-01
5.68555474e-01 2.39237025e-01 -8.70676994e-01 2.29313150e-01
1.34502447e+00 3.76657486e-01 4.09246862e-01 5.88212967e-01
7.81266510e-01 2.62988210e-01 1.06193650e+00 3.62749696e-01
1.41554847e-01 9.84309077e-01 6.12092555e-01 -2.62393560e-02
-6.41090512e-01 2.20399171e-01 4.56216782e-01 4.75740582e-01
-4.06262577e-01 -1.39908671e-01 -7.56178379e-01 7.38828361e-01
-2.35167909e+00 -1.19228721e+00 -7.57708311e-01 2.39897633e+00
2.13589981e-01 2.83028096e-01 7.04287231e-01 -5.38398623e-02
1.25321150e+00 -7.89808705e-02 -2.87148714e-01 1.47273660e-01
-1.32398933e-01 -3.89363579e-02 9.70975161e-01 4.39442307e-01
-1.33212733e+00 6.79084241e-01 6.72122097e+00 7.38989830e-01
-8.14374089e-01 4.66245294e-01 -4.49568182e-01 -6.53972700e-02
5.94181776e-01 1.31682619e-01 -1.53424299e+00 4.75335479e-01
1.08229256e+00 -1.71098888e-01 -1.57618880e-01 4.82512385e-01
1.00349821e-01 -2.68873155e-01 -1.13753474e+00 1.03129303e+00
1.12732783e-01 -1.39847350e+00 -2.63070703e-01 1.27210110e-01
4.07225043e-01 -5.82585149e-02 -6.19558282e-02 5.24007529e-02
1.17230535e-01 -4.75106657e-01 1.00939119e+00 8.47364902e-01
2.24794105e-01 -2.15200603e-01 5.90722501e-01 2.22076289e-03
-1.83292699e+00 -8.15396681e-02 -8.78031403e-02 1.30455658e-01
4.25513059e-01 4.06819046e-01 -7.81806290e-01 7.94983149e-01
8.40417206e-01 8.61401856e-01 -9.13561106e-01 1.91955423e+00
4.14354205e-01 2.98258215e-01 -4.90057617e-01 8.47298056e-02
8.36711526e-02 3.39621715e-02 1.24430978e+00 1.24105871e+00
4.64989036e-01 -2.86716014e-01 2.98195869e-01 6.51031733e-01
5.81580460e-01 -2.96955287e-01 -4.93070781e-01 2.06091434e-01
2.54269838e-01 1.42464364e+00 -1.07330322e+00 -5.85599005e-01
-3.17180336e-01 4.60218608e-01 -1.92338809e-01 -8.02744739e-03
-1.18294597e+00 -1.46164447e-01 4.29769546e-01 4.10541445e-01
6.28377140e-01 -3.59946847e-01 -4.78454214e-03 -6.45808995e-01
-1.31252602e-01 -3.37807000e-01 7.04332113e-01 -5.88843584e-01
-9.35713053e-01 5.31716585e-01 2.83533007e-01 -1.79701781e+00
-8.50253701e-02 -5.62317848e-01 -4.39052224e-01 4.82607484e-01
-1.04070032e+00 -1.39423001e+00 -3.39534700e-01 6.56943083e-01
2.98910767e-01 -2.43010491e-01 3.69525820e-01 9.52186644e-01
-3.08280438e-01 3.91978860e-01 -1.21959999e-01 -3.90885584e-02
8.24547768e-01 -1.21541166e+00 -1.02921631e-02 1.10600233e+00
2.60584086e-01 3.81437063e-01 1.05325186e+00 -1.03494787e+00
-1.09527898e+00 -1.18731129e+00 4.61948633e-01 -9.69720185e-01
6.82286143e-01 -1.41229436e-01 -1.02319372e+00 9.71667528e-01
2.88043886e-01 1.18403144e-01 3.83401811e-01 -4.69947122e-02
1.11178912e-01 -2.07976162e-01 -1.00209153e+00 -5.14509678e-02
1.24799705e+00 -3.58599909e-02 -6.67270780e-01 2.52992183e-01
3.78768504e-01 -4.63736176e-01 -9.00566399e-01 6.26646698e-01
7.45365083e-01 -9.30354059e-01 9.43843842e-01 -1.56981722e-01
-8.82792532e-01 -1.03993511e+00 2.21126810e-01 -6.47722661e-01
-4.28346872e-01 -7.33729839e-01 -3.87177199e-01 1.34695375e+00
8.99026841e-02 -5.74228346e-01 9.47307348e-01 1.63292050e-01
-7.96882585e-02 2.12574750e-02 -1.14088500e+00 -1.47161818e+00
-5.41629851e-01 -1.65141270e-01 2.87390828e-01 7.79154062e-01
-5.22214115e-01 -4.07650359e-02 -5.86894691e-01 3.89638871e-01
1.12936473e+00 1.91916570e-01 8.92155170e-01 -1.76380372e+00
-2.75831223e-01 -2.10356474e-01 -8.43456268e-01 -7.92605877e-01
-3.47672313e-01 -6.61929309e-01 4.43928316e-03 -1.18054640e+00
1.60027683e-01 -6.17322326e-01 -4.81097579e-01 3.96031618e-01
9.76803601e-02 4.90476251e-01 6.30742073e-01 2.82026529e-01
-1.33335578e+00 4.80737276e-02 9.93908405e-01 5.35113253e-02
-1.41912639e-01 1.50541529e-01 1.23230547e-01 5.73300600e-01
4.36438262e-01 -9.12435889e-01 2.36324072e-01 -9.53935981e-02
1.10423192e-02 -9.87977684e-02 7.70348608e-01 -1.83366430e+00
7.51392841e-01 3.22549611e-01 2.93082535e-01 -1.26983500e+00
5.17016768e-01 -1.12030709e+00 8.69133532e-01 9.19712543e-01
1.40707418e-01 3.98832202e-01 5.88993371e-01 6.04182720e-01
1.76267070e-03 -3.42743069e-01 8.54746878e-01 -3.12539414e-02
-1.13852131e+00 3.57694149e-01 -3.74625653e-01 -4.68495935e-01
1.51005304e+00 -6.83647394e-01 -4.54416722e-01 1.35510579e-01
-1.16211367e+00 4.17384207e-01 3.59495997e-01 5.81268549e-01
2.32664242e-01 -1.49263024e+00 -5.25590539e-01 -4.65686992e-02
-5.73287420e-02 -5.56411862e-01 3.32945645e-01 1.30978775e+00
-2.30314046e-01 4.28233832e-01 -6.39142752e-01 -1.19703472e+00
-1.84129429e+00 5.85782707e-01 3.81348580e-01 -2.50016719e-01
-7.10179210e-01 4.22889501e-01 -1.48464873e-01 7.83128366e-02
2.99998224e-01 -1.66032359e-01 -3.90756249e-01 3.01746130e-01
5.70372164e-01 8.97728801e-01 -3.24851349e-02 -7.26082325e-01
-5.80941260e-01 9.65898275e-01 8.84026289e-03 -6.40017614e-02
9.78232503e-01 -3.14070910e-01 6.15721717e-02 5.33392608e-01
4.23119098e-01 -1.20080441e-01 -1.33947062e+00 -1.01997986e-01
3.07690710e-01 -6.44073665e-01 -3.27109039e-01 -5.89549601e-01
-9.42943633e-01 5.68472505e-01 1.50070572e+00 4.61279750e-01
5.91960192e-01 7.03273481e-03 6.03709757e-01 1.45691827e-01
4.49459940e-01 -1.01627648e+00 1.60387889e-01 4.38764691e-01
4.35085416e-01 -1.10871005e+00 2.07169391e-02 -2.83230066e-01
-8.83118510e-02 1.07359958e+00 8.86774957e-01 -1.04307242e-01
5.71709633e-01 4.94841009e-01 1.63106680e-01 -3.07349384e-01
-3.27879578e-01 -7.23439038e-01 2.00948671e-01 8.19509089e-01
-6.90897033e-02 -1.36662215e-01 -3.10619503e-01 -9.92079377e-02
1.98736444e-01 1.36192858e-01 4.35721755e-01 9.52335358e-01
-6.78352475e-01 -1.18186462e+00 -9.72361088e-01 2.72549063e-01
-2.80025899e-01 4.91967142e-01 2.06896830e-02 1.06754684e+00
4.72204745e-01 8.73666763e-01 2.50078380e-01 -3.31253916e-01
3.41316432e-01 -1.52118355e-01 7.52534270e-01 -4.57075447e-01
-8.74628544e-01 1.11019470e-01 -7.39358068e-02 -8.11877608e-01
-1.17159009e+00 -9.90961254e-01 -1.22128868e+00 -2.50886500e-01
-7.33118832e-01 2.60683060e-01 6.82279885e-01 9.09267545e-01
3.47253382e-01 7.03458846e-01 1.12537049e-01 -9.59598780e-01
1.41326323e-01 -7.95945406e-01 -5.85389197e-01 3.48668963e-01
5.08480966e-01 -1.18481648e+00 -7.21923187e-02 8.72662738e-02] | [6.514688014984131, -2.0205962657928467] |
60b05cec-2056-42a5-b19b-2be06673c2ce | knowledge-guided-paraphrase-identification | null | null | https://aclanthology.org/2021.findings-emnlp.72 | https://aclanthology.org/2021.findings-emnlp.72.pdf | Knowledge-Guided Paraphrase Identification | Paraphrase identification (PI), a fundamental task in natural language processing, is to identify whether two sentences express the same or similar meaning, which is a binary classification problem. Recently, BERT-like pre-trained language models have been a popular choice for the frameworks of various PI models, but almost all existing methods consider general domain text. When these approaches are applied to a specific domain, existing models cannot make accurate predictions due to the lack of professional knowledge. In light of this challenge, we propose a novel framework, namely , which can leverage the external unstructured Wikipedia knowledge to accurately identify paraphrases. We propose to mine outline knowledge of concepts related to given sentences from Wikipedia via BM25 model. After retrieving related outline knowledge, makes predictions based on both the semantic information of two sentences and the outline knowledge. Besides, we propose a gating mechanism to aggregate the semantic information-based prediction and the knowledge-based prediction. Extensive experiments are conducted on two public datasets: PARADE (a computer science domain dataset) and clinicalSTS2019 (a biomedical domain dataset). The results show that the proposed outperforms state-of-the-art methods. | ['Jing Gao', 'Yaqing Wang', 'Fenglong Ma', 'Haoyu Wang'] | null | null | null | null | findings-emnlp-2021-11 | ['paraphrase-identification'] | ['natural-language-processing'] | [ 4.37060118e-01 -3.71067040e-02 -5.12085259e-01 -3.98991585e-01
-8.11851680e-01 -3.41239303e-01 3.36315691e-01 6.30832374e-01
-4.88127381e-01 7.69831181e-01 4.97908026e-01 -1.99526981e-01
-3.67190242e-02 -7.82803237e-01 -6.23027623e-01 -1.85530111e-01
5.75874805e-01 4.19490576e-01 4.17627275e-01 -2.03083187e-01
6.20768011e-01 -1.88825846e-01 -1.09370923e+00 1.04809606e+00
1.05488944e+00 9.00226355e-01 4.54182714e-01 1.07819960e-01
-6.18794322e-01 9.00748849e-01 -3.23516399e-01 -6.28771424e-01
-2.28760779e-01 -4.63132679e-01 -1.17115545e+00 -2.71194845e-01
-1.59418747e-01 1.60163239e-01 -2.30356574e-01 1.22403789e+00
4.69615489e-01 -1.41681731e-01 4.74965900e-01 -9.32197392e-01
-7.92105019e-01 5.13005197e-01 -4.45689768e-01 3.83929461e-01
7.25948095e-01 -1.85597464e-01 1.17269802e+00 -9.11003947e-01
8.14514816e-01 1.12713075e+00 6.70386672e-01 4.84681904e-01
-9.51882839e-01 -4.84296143e-01 6.44763559e-02 7.54273534e-01
-1.22805619e+00 -3.07375729e-01 9.60024595e-01 -2.02350765e-01
1.03740287e+00 4.37320508e-02 3.74743342e-01 1.18311584e+00
3.30699980e-01 1.14592123e+00 1.02137864e+00 -4.32079613e-01
1.47388801e-01 3.83179694e-01 4.87626314e-01 5.24423957e-01
1.39379978e-01 -4.89581347e-01 -6.03563964e-01 -4.86242861e-01
4.84943986e-02 3.49551678e-01 -5.66945553e-01 -5.45291975e-02
-1.19661891e+00 6.90841198e-01 3.73831958e-01 3.94211292e-01
-3.60804319e-01 -5.48573315e-01 6.82387710e-01 4.70116138e-01
3.63142848e-01 5.29282510e-01 -5.15942633e-01 -5.99069335e-02
-8.78847599e-01 6.98093399e-02 1.00014234e+00 9.76526976e-01
7.34386206e-01 -1.03880227e+00 -3.20347071e-01 1.09705389e+00
2.70425856e-01 2.24019915e-01 1.12707579e+00 -3.07787091e-01
8.55292201e-01 1.12381303e+00 -1.23156719e-01 -1.22385991e+00
-3.65269095e-01 -3.72285932e-01 -8.37270439e-01 -8.00908089e-01
1.48609504e-01 5.53130433e-02 -5.66420138e-01 1.55276453e+00
2.46732496e-02 3.67862552e-01 3.98463011e-01 6.23604357e-01
1.38622010e+00 4.22759801e-01 2.25009620e-01 -2.82237947e-01
1.62166452e+00 -1.04400110e+00 -7.26704061e-01 -4.87548113e-01
6.33042514e-01 -8.13973725e-01 9.47141528e-01 1.85729459e-01
-7.48857379e-01 -5.49303293e-01 -8.86827528e-01 -2.33433977e-01
-3.51365864e-01 3.05842310e-01 2.96866477e-01 2.06148863e-01
-7.26069450e-01 6.28575802e-01 -4.97223049e-01 -7.17949986e-01
4.62126106e-01 -3.89921330e-02 -3.43983710e-01 -3.87330323e-01
-1.59055674e+00 9.25598621e-01 5.84578335e-01 2.54143551e-02
-3.80424142e-01 -6.31727397e-01 -6.58243477e-01 2.55387813e-01
2.19827905e-01 -1.09997821e+00 1.00227284e+00 -1.05240715e+00
-1.23582256e+00 1.33722663e+00 -4.85533476e-01 -5.52292049e-01
3.66840005e-01 -5.87440096e-02 -4.61525261e-01 3.87126684e-01
4.20360595e-01 3.29948753e-01 6.27285957e-01 -7.11678505e-01
-5.91942012e-01 -5.26441276e-01 1.58958778e-01 3.92143816e-01
-5.36259592e-01 2.08424982e-02 -5.85761845e-01 -4.58652318e-01
2.09497020e-01 -7.75096834e-01 -1.58476084e-01 -3.62554975e-02
-6.75091684e-01 -4.86214638e-01 2.67794251e-01 -8.99943888e-01
1.42291915e+00 -2.09515691e+00 8.10970590e-02 9.54444520e-03
1.99400499e-01 2.71913350e-01 1.89142507e-02 5.82437634e-01
-2.65033022e-02 -1.81620829e-02 -3.28880072e-01 -1.95614889e-01
-2.06775934e-01 6.29113317e-02 -3.94434094e-01 -7.29214922e-02
1.01473898e-01 8.97046804e-01 -1.16301680e+00 -9.33791578e-01
-1.88195422e-01 -1.97294623e-01 -4.44898576e-01 2.97930092e-01
-9.25245360e-02 2.53224283e-01 -8.42945158e-01 5.56701958e-01
6.06590569e-01 -5.48142910e-01 3.22184682e-01 -3.35513204e-01
4.55989003e-01 4.86296296e-01 -5.85795045e-01 2.08586740e+00
-4.68946457e-01 1.27448022e-01 -2.63845831e-01 -1.52241373e+00
8.12672079e-01 3.60615343e-01 2.47793779e-01 -5.61585367e-01
-1.36116162e-01 3.83157015e-01 -1.53332204e-01 -9.45647895e-01
5.14659323e-02 -4.47424829e-01 -2.61112954e-02 2.73517489e-01
-5.30914180e-02 4.32216488e-02 2.40352228e-01 2.64384598e-01
1.37926698e+00 -1.35619357e-01 6.96893156e-01 -2.82036394e-01
1.04800963e+00 1.44205660e-01 7.62393355e-01 7.87689984e-01
-3.28464538e-01 5.15275359e-01 5.49943924e-01 -4.37224567e-01
-5.91507494e-01 -8.57794881e-01 -1.94217399e-01 7.38081276e-01
4.10762995e-01 -4.92912710e-01 -5.43932438e-01 -9.65088725e-01
6.51314035e-02 6.53447688e-01 -6.04126871e-01 -5.02052069e-01
-3.10488969e-01 -9.05895233e-01 4.26423639e-01 4.20706928e-01
6.13903821e-01 -1.22318304e+00 -1.12215430e-01 3.75806361e-01
-8.06729257e-01 -1.29139161e+00 -3.86899292e-01 -1.64842799e-01
-9.36099291e-01 -1.30481410e+00 -7.35147238e-01 -1.08680415e+00
6.45357668e-01 2.56921560e-01 1.22866881e+00 4.84257974e-02
-1.61576286e-01 8.29625428e-02 -5.79097390e-01 -4.17613208e-01
-4.04720396e-01 2.03768834e-01 -2.69080251e-01 5.48411570e-02
9.71934855e-01 -5.18253684e-01 -8.04137409e-01 1.43970728e-01
-7.05415249e-01 3.42645973e-01 9.21192348e-01 1.02079058e+00
6.63749754e-01 -2.79063463e-01 9.62946951e-01 -1.05891252e+00
8.97586226e-01 -9.19068456e-01 2.26063073e-01 6.76991463e-01
-4.10127848e-01 8.07879865e-02 8.84595156e-01 -1.92691103e-01
-1.08835399e+00 -1.58738613e-01 -3.41343820e-01 -1.71082929e-01
-1.06451519e-01 1.01582479e+00 -8.12719017e-02 3.34280252e-01
5.16356111e-01 8.19361866e-01 2.53972933e-02 -5.67457795e-01
4.37736996e-02 1.10973096e+00 3.34980458e-01 -4.37145740e-01
2.65647054e-01 4.17489082e-01 -3.75200719e-01 -6.22387886e-01
-1.41765594e+00 -8.78980100e-01 -3.62941593e-01 3.25174212e-01
7.82725275e-01 -9.81876612e-01 -3.30216169e-01 2.29846090e-01
-1.26075423e+00 2.83348173e-01 1.44599125e-01 3.12463731e-01
-5.14172554e-01 9.45177436e-01 -5.79503894e-01 -1.89144284e-01
-8.17586601e-01 -9.90662992e-01 9.30906832e-01 2.57396221e-01
-3.23951632e-01 -9.81704772e-01 1.54360563e-01 8.57547641e-01
1.71259016e-01 -2.94459879e-01 1.20260990e+00 -9.85791624e-01
-1.03422150e-01 -2.38824993e-01 -4.13533986e-01 2.29123309e-01
2.84574509e-01 -4.92751598e-01 -6.99342728e-01 -1.22683018e-01
2.20436946e-01 -5.08607447e-01 1.08408093e+00 6.61196858e-02
1.49476302e+00 -2.41200656e-01 -7.22696185e-01 3.18832457e-01
1.21164978e+00 -1.22610658e-01 5.26484668e-01 4.19229835e-01
3.62473458e-01 6.28431320e-01 6.97977722e-01 4.78071511e-01
7.00265884e-01 5.92251182e-01 -5.98207265e-02 2.63960093e-01
-4.62131612e-02 -4.97698188e-01 1.65460184e-01 1.01491308e+00
3.91344130e-01 -3.84390280e-02 -1.00391638e+00 6.47977054e-01
-2.08401251e+00 -1.04448581e+00 9.27444845e-02 1.94213974e+00
1.31798708e+00 3.18836659e-01 -3.74952376e-01 -1.02713436e-01
7.09508538e-01 -7.54691586e-02 -6.25511050e-01 -1.86099023e-01
-4.87581454e-02 6.32282868e-02 2.41391137e-02 1.83376670e-02
-1.03554332e+00 7.72665381e-01 5.08425808e+00 1.18603730e+00
-8.65964174e-01 1.57659158e-01 6.21875048e-01 3.47301930e-01
-2.98969060e-01 -5.49454577e-02 -8.01665723e-01 8.91057849e-01
8.28471899e-01 -5.87139368e-01 -6.29872307e-02 8.83625329e-01
2.49953434e-01 1.59406383e-02 -1.25399089e+00 1.10078204e+00
4.24016267e-01 -1.38347411e+00 3.62327456e-01 -5.24137199e-01
3.86339068e-01 9.69665945e-02 -2.85298884e-01 5.84936440e-01
-8.62063318e-02 -7.86295831e-01 7.84850568e-02 7.70272911e-01
3.83588165e-01 -2.92373538e-01 1.14950728e+00 6.94360733e-01
-1.08635771e+00 -1.48369923e-01 -7.11382508e-01 -6.66675158e-04
-2.21950840e-02 7.54265428e-01 -8.34056139e-01 8.17263663e-01
6.08309984e-01 1.16495657e+00 -6.46671057e-01 1.23057437e+00
-4.58021075e-01 6.38773799e-01 -1.36968806e-01 -3.25666398e-01
7.47983977e-02 -2.86454167e-02 3.28868926e-01 1.28999531e+00
1.44865245e-01 8.67067203e-02 2.73234755e-01 8.51871848e-01
-3.30274969e-01 6.42799258e-01 -4.82631624e-01 -4.53674980e-02
4.65398163e-01 1.10305274e+00 -4.97518420e-01 -5.93047798e-01
-7.72267342e-01 1.14346194e+00 5.55692554e-01 1.08775310e-01
-5.58530629e-01 -6.30440235e-01 3.34874421e-01 -1.27113014e-01
2.57026926e-02 4.87501591e-01 -2.31735200e-01 -1.56727803e+00
4.00622010e-01 -8.00617397e-01 6.77798271e-01 -6.78260446e-01
-1.85740888e+00 4.67586905e-01 -2.49003723e-01 -1.49053276e+00
6.00605682e-02 -4.91977602e-01 -7.57734418e-01 6.97015226e-01
-1.86137915e+00 -1.17834866e+00 -2.80846775e-01 5.55676937e-01
8.00503552e-01 -1.87442049e-01 8.95748138e-01 3.14882278e-01
-5.91995537e-01 5.46978235e-01 3.43950987e-01 3.56340528e-01
1.05973518e+00 -1.02040768e+00 1.62874475e-01 4.65362400e-01
3.17971483e-02 9.30607736e-01 5.50121546e-01 -6.98192954e-01
-1.11446655e+00 -1.03126025e+00 1.50572681e+00 -3.92167002e-01
7.07043707e-01 7.35317394e-02 -1.07110739e+00 3.70955944e-01
5.09392144e-03 -7.70959482e-02 1.16564977e+00 4.45398241e-02
-2.06029341e-01 -5.27160466e-02 -1.05094039e+00 3.49022657e-01
1.12173402e+00 -7.57406056e-01 -1.32191825e+00 5.87363124e-01
7.36872971e-01 -1.68061778e-01 -7.74992168e-01 3.60833764e-01
4.51197028e-01 -7.33104408e-01 9.63296235e-01 -8.91261220e-01
9.22266603e-01 -2.26557150e-01 -2.24953294e-02 -1.29427862e+00
-1.88672438e-01 3.12499539e-03 1.67139307e-01 1.02896082e+00
5.91686904e-01 -5.50407231e-01 7.65319169e-01 5.55743694e-01
-2.31710032e-01 -9.00367558e-01 -9.45499420e-01 -5.18411100e-01
2.29580943e-02 -1.05954327e-01 4.11643565e-01 1.07140720e+00
7.71296263e-01 6.43576801e-01 -7.41325021e-02 -3.46501134e-02
2.44793579e-01 6.96927071e-01 3.42678934e-01 -1.06123984e+00
-4.24096227e-01 -2.88904160e-01 -5.36554694e-01 -1.10864127e+00
5.44692993e-01 -1.35844123e+00 -1.11879714e-01 -1.82419205e+00
1.04455161e+00 -1.92962363e-01 -6.34833872e-01 4.74278212e-01
-6.80889606e-01 -7.22334310e-02 -6.61209673e-02 4.93715256e-01
-9.97305453e-01 5.98038256e-01 9.65784431e-01 -3.87039512e-01
-2.92860530e-02 2.59471089e-01 -8.63627315e-01 8.51264894e-01
6.70921385e-01 -4.64066207e-01 -6.03789806e-01 -4.07739192e-01
3.66263449e-01 1.36677921e-01 2.08351359e-01 -9.68993187e-01
4.24203098e-01 -5.66528626e-02 1.00645989e-01 -4.29855078e-01
-3.84545848e-02 -5.93457282e-01 -2.63865143e-01 5.86306572e-01
-7.12593615e-01 -5.18507585e-02 7.66792968e-02 9.02468443e-01
-5.96711636e-01 -5.58983386e-01 5.92496872e-01 -3.89609069e-01
-8.34643245e-01 2.46062353e-01 -1.07770555e-01 5.18016219e-01
8.45214427e-01 6.89451844e-02 -3.07253778e-01 -1.34427711e-01
-7.54453361e-01 4.76106822e-01 2.16388628e-01 5.48439562e-01
8.11389923e-01 -1.11005545e+00 -7.92002261e-01 -8.72032195e-02
6.16006851e-01 -2.45531738e-01 5.04153550e-01 8.42815936e-01
-2.78705776e-01 5.32700539e-01 -1.82687625e-01 -3.72406423e-01
-1.28381264e+00 7.02309370e-01 6.93620667e-02 -7.87232041e-01
-4.99800593e-01 8.13334882e-01 3.50655824e-01 -5.08118212e-01
-1.19989865e-01 -2.79580563e-01 -6.25964880e-01 4.78598624e-02
7.02126980e-01 -2.09952787e-01 2.41981447e-01 -3.33830684e-01
-5.95853627e-01 4.01184976e-01 -4.92624491e-01 4.44143862e-01
1.16566563e+00 -2.38043919e-01 -3.34940314e-01 2.59341270e-01
1.23398495e+00 -4.21049595e-01 -3.31646383e-01 -8.02011669e-01
3.21671665e-01 -2.64181793e-01 -1.86101004e-01 -8.33960354e-01
-6.45440698e-01 1.05448902e+00 2.51760602e-01 -2.38750771e-01
1.03166628e+00 7.83350691e-02 1.04152668e+00 7.47946441e-01
4.56857115e-01 -1.14820015e+00 1.83163568e-01 4.65870112e-01
8.59197259e-01 -1.46038508e+00 -9.76318792e-02 -5.85674167e-01
-7.02903509e-01 1.08568096e+00 6.11876547e-01 1.02633037e-01
6.63224041e-01 -4.24905241e-01 -1.28053471e-01 -2.67748684e-02
-8.27972949e-01 -2.15919558e-02 3.15965742e-01 3.23521852e-01
5.43010652e-01 -1.10024922e-01 -7.70373702e-01 1.23899531e+00
-8.14077258e-03 4.54818487e-01 2.32836246e-01 8.65978777e-01
-4.11111921e-01 -1.21131730e+00 1.49821296e-01 9.78527009e-01
-6.41774714e-01 -4.57373291e-01 -3.90754223e-01 6.35261238e-02
5.41922599e-02 1.04693174e+00 -3.82950544e-01 -2.46101603e-01
3.19857538e-01 2.64295787e-01 1.06492728e-01 -9.25418198e-01
-5.18844843e-01 -6.00592613e-01 2.14667648e-01 -4.08665150e-01
-4.41198140e-01 -4.47402298e-01 -1.19613326e+00 9.43267867e-02
-1.40514582e-01 3.89546275e-01 1.16588540e-01 1.17775345e+00
6.91867113e-01 1.98455244e-01 4.26161528e-01 -1.61685407e-01
-8.24004471e-01 -9.66444552e-01 -3.66714537e-01 1.01571190e+00
-7.79882669e-02 -4.53191549e-01 -1.38739690e-01 5.27572073e-02] | [10.995865821838379, 8.440045356750488] |
32115980-47d5-430f-8229-04f7cb5c42df | boosting-camouflaged-object-detection-with | 2205.10579 | null | https://arxiv.org/abs/2205.10579v1 | https://arxiv.org/pdf/2205.10579v1.pdf | Boosting Camouflaged Object Detection with Dual-Task Interactive Transformer | Camouflaged object detection intends to discover the concealed objects hidden in the surroundings. Existing methods follow the bio-inspired framework, which first locates the object and second refines the boundary. We argue that the discovery of camouflaged objects depends on the recurrent search for the object and the boundary. The recurrent processing makes the human tired and helpless, but it is just the advantage of the transformer with global search ability. Therefore, a dual-task interactive transformer is proposed to detect both accurate position of the camouflaged object and its detailed boundary. The boundary feature is considered as Query to improve the camouflaged object detection, and meanwhile the object feature is considered as Query to improve the boundary detection. The camouflaged object detection and the boundary detection are fully interacted by multi-head self-attention. Besides, to obtain the initial object feature and boundary feature, transformer-based backbones are adopted to extract the foreground and background. The foreground is just object, while foreground minus background is considered as boundary. Here, the boundary feature can be obtained from blurry boundary region of the foreground and background. Supervised by the object, the background and the boundary ground truth, the proposed model achieves state-of-the-art performance in public datasets. https://github.com/liuzywen/COD | ['Wei Wu', 'Zhili Zhang', 'Zhengyi Liu'] | 2022-05-21 | null | null | null | null | ['boundary-detection'] | ['computer-vision'] | [ 2.86334038e-01 -1.87973112e-01 -2.44924799e-02 1.19231403e-01
-5.22666633e-01 -5.42455196e-01 3.81728470e-01 -4.30289477e-01
-2.19970435e-01 5.48259735e-01 1.62050985e-02 4.62497286e-02
1.33827686e-01 -5.76176405e-01 -4.49362546e-01 -1.28812337e+00
3.78860354e-01 2.99162894e-01 8.35720718e-01 8.67548212e-02
4.50541794e-01 3.10188264e-01 -1.70887506e+00 2.67805517e-01
7.88861334e-01 1.35300767e+00 6.79178298e-01 4.46153253e-01
-3.22481334e-01 4.71507758e-01 -7.24367023e-01 -4.02729362e-02
2.63790041e-01 -3.20811808e-01 -5.39980590e-01 2.88863152e-01
3.28849822e-01 -4.69305426e-01 -1.70575112e-01 1.36222935e+00
3.49518269e-01 -3.82062979e-02 3.87494266e-01 -1.29579473e+00
-7.23036230e-01 -1.69191379e-02 -8.70704651e-01 6.09643281e-01
8.28113183e-02 3.66626918e-01 6.81299627e-01 -1.18551910e+00
5.07828176e-01 1.19134712e+00 8.33153129e-02 5.33431649e-01
-6.31412983e-01 -9.05258358e-01 5.29247105e-01 5.93675494e-01
-1.65013766e+00 -4.89151597e-01 1.02262819e+00 -2.85510629e-01
1.70742482e-01 5.56829154e-01 8.40046823e-01 6.43699348e-01
1.87123995e-02 1.23805034e+00 7.34621584e-01 -1.24701865e-01
-9.83920246e-02 1.71790034e-01 1.61368027e-01 8.31888497e-01
3.04993927e-01 1.07169390e-01 -1.85507715e-01 -4.23283465e-02
5.64651072e-01 3.39057535e-01 -7.39814758e-01 -1.84692845e-01
-1.18809974e+00 3.39975774e-01 4.14180368e-01 3.67749423e-01
-1.70567453e-01 -8.63403156e-02 2.26775125e-01 -2.46222436e-01
4.45255637e-01 1.06232308e-01 -4.74683851e-01 3.83188426e-02
-9.89991665e-01 6.34379163e-02 3.12803924e-01 8.50350559e-01
7.84660459e-01 -3.37794274e-01 -4.71900463e-01 3.43109787e-01
2.60615885e-01 6.99731588e-01 2.51891702e-01 -7.33390510e-01
6.63586184e-02 8.37510943e-01 5.18152475e-01 -1.17516983e+00
7.47249797e-02 -3.43192756e-01 -6.80134356e-01 1.58393160e-01
3.44704688e-01 1.61823362e-01 -1.12028873e+00 1.32878208e+00
9.30646658e-01 6.12522185e-01 -3.50503474e-01 1.37304425e+00
1.08648896e+00 7.59196758e-01 -2.28153244e-01 -4.78045344e-01
1.81077552e+00 -1.27357781e+00 -1.00542736e+00 -2.43895382e-01
3.87307376e-01 -9.60544765e-01 9.70716000e-01 4.42273170e-01
-9.05469358e-01 -5.87720275e-01 -6.94434881e-01 -1.28049940e-01
-3.32769841e-01 4.45055604e-01 4.13096726e-01 3.13515246e-01
-6.65586829e-01 -2.94513255e-02 -4.72676963e-01 -5.45342676e-02
7.05578625e-01 2.46409550e-01 -8.45283791e-02 -1.98631912e-01
-1.03389704e+00 7.93648601e-01 4.52809691e-01 6.23305142e-01
-1.06938493e+00 -4.53720272e-01 -5.33769667e-01 2.06283741e-02
8.71752083e-01 -5.26688099e-01 8.74656081e-01 -1.06308782e+00
-1.02836180e+00 1.04480207e+00 -4.65447336e-01 -1.14485705e-02
6.79870129e-01 -1.11943975e-01 -3.90924543e-01 2.58012891e-01
2.38914937e-01 5.23020148e-01 1.26692700e+00 -1.28969443e+00
-1.24120235e+00 -2.64350891e-01 -3.56012374e-01 2.55882084e-01
-1.00069642e-01 -7.55794197e-02 -9.62210774e-01 -5.29889464e-01
3.30451965e-01 -5.59163272e-01 2.53785282e-01 2.20030889e-01
-5.22133172e-01 -5.11717618e-01 1.78352702e+00 -8.52398992e-01
1.28867495e+00 -2.39131951e+00 -1.21995509e-01 -1.82335675e-01
6.01096749e-01 5.02309263e-01 3.16030420e-02 -2.99175233e-01
1.99131668e-01 1.07453233e-02 -1.23320231e-02 -9.80148390e-02
-4.16685522e-01 -7.33774006e-02 -4.13293630e-01 8.16281617e-01
2.17605770e-01 9.72369611e-01 -8.40276718e-01 -8.35767567e-01
3.02509457e-01 3.96143436e-01 -1.06245123e-01 3.47692758e-01
-2.14445591e-01 7.25286305e-01 -6.94589019e-01 9.70033705e-01
1.12363672e+00 -3.52758944e-01 -5.45329154e-01 -4.78572130e-01
-2.62781233e-01 -1.30399182e-01 -1.27418888e+00 1.09134007e+00
1.44868404e-01 4.36119020e-01 5.51891744e-01 -6.44450486e-01
9.08825219e-01 1.66164972e-02 3.13849151e-01 -6.85022473e-01
3.90526086e-01 2.84155041e-01 -5.21876030e-02 -8.87610972e-01
2.20028892e-01 1.98944226e-01 3.17707092e-01 1.73848808e-01
-5.31199336e-01 1.63873643e-01 -9.00744200e-02 8.41707438e-02
8.12352240e-01 2.52460033e-01 1.78862847e-02 -2.44752660e-01
8.45063388e-01 1.12469895e-02 8.07538629e-01 4.88082945e-01
-4.47744429e-01 5.56948245e-01 2.66102497e-02 -5.15091240e-01
-5.34449041e-01 -9.50406194e-01 -1.33588687e-01 1.00184906e+00
1.23542297e+00 1.58279911e-01 -7.41915166e-01 -8.03761303e-01
-5.46245426e-02 4.75780547e-01 -8.27158570e-01 -3.01737636e-01
-6.88842773e-01 -4.19146717e-01 1.18384100e-01 1.00319058e-01
9.91726995e-01 -1.18250656e+00 -9.09173489e-01 -6.74603507e-02
-6.30153179e-01 -7.95769989e-01 -8.13053071e-01 -2.46701509e-01
-4.44266558e-01 -1.26241672e+00 -9.64895427e-01 -8.33083570e-01
7.07834482e-01 8.70938957e-01 5.68948269e-01 7.93336451e-01
-4.64022964e-01 -6.71316823e-03 -1.05452783e-01 -2.60112017e-01
1.06502905e-01 -4.01392311e-01 -9.45205092e-02 4.70360279e-01
1.99084058e-01 -1.17518812e-01 -9.00504470e-01 6.38230503e-01
-6.32449567e-01 2.62265027e-01 5.14308751e-01 7.33368933e-01
6.65367424e-01 3.88939917e-01 1.99181497e-01 -2.99956739e-01
1.10483408e-01 -3.17403257e-01 -7.83917785e-01 2.62546957e-01
-3.52388680e-01 -3.89170229e-01 1.81139320e-01 -8.62950385e-01
-1.18883955e+00 -1.16596289e-01 2.65107572e-01 -8.17532897e-01
-2.21612737e-01 -9.05122459e-02 -6.26529157e-01 -1.30535495e-02
8.20708349e-02 5.47502220e-01 -3.18403780e-01 -6.29214704e-01
1.90131709e-01 5.50061941e-01 7.07057714e-01 -1.62413508e-01
7.22247958e-01 8.16523552e-01 -2.86198080e-01 -6.89987957e-01
-8.39635134e-01 -6.85428798e-01 -4.15975899e-01 -4.18118149e-01
9.26596284e-01 -5.82938492e-01 -1.03558707e+00 6.16160452e-01
-1.44969106e+00 -1.22095183e-01 -3.27972211e-02 1.26088098e-01
5.57657778e-02 4.62625444e-01 -4.31441694e-01 -1.15422630e+00
-5.57574868e-01 -1.11369324e+00 1.48546171e+00 5.90405881e-01
3.95463258e-01 -5.64881742e-01 -4.79983449e-01 4.52232897e-01
3.93796444e-01 2.68726144e-02 5.47338963e-01 -4.05510008e-01
-1.21484268e+00 -1.97821170e-01 -7.50099659e-01 -2.20136438e-02
3.19234461e-01 -8.93704817e-02 -1.06203079e+00 -1.71164095e-01
3.35699350e-01 2.90475219e-01 9.75379050e-01 4.95259762e-01
1.15783620e+00 -4.39101309e-01 -9.56169903e-01 6.77486718e-01
1.08626270e+00 4.13536221e-01 6.83602571e-01 2.90726632e-01
8.70079458e-01 4.32142407e-01 1.17028499e+00 2.06529140e-01
1.37702838e-01 4.43438590e-01 7.03434527e-01 -3.43617469e-01
-3.49353552e-01 -5.89776039e-02 2.64689237e-01 2.81863660e-01
2.20466897e-01 -4.37936425e-01 -6.84001148e-01 6.44227922e-01
-1.74165893e+00 -9.61693943e-01 -2.86902010e-01 2.08337855e+00
6.39928937e-01 2.08841622e-01 -1.26406595e-01 -2.68629845e-02
1.30501056e+00 9.71410722e-02 -8.00188601e-01 4.46809471e-01
-3.44596088e-01 -4.39007759e-01 3.77988309e-01 4.61165518e-01
-1.15181220e+00 1.16939950e+00 4.95219374e+00 1.22521663e+00
-1.39144933e+00 2.66442120e-01 6.70776606e-01 -1.01666108e-01
-2.97330134e-02 7.39170387e-02 -1.07016289e+00 7.90035665e-01
-1.24292918e-01 -5.37361577e-03 4.67619419e-01 8.15527737e-01
2.43187472e-01 -6.39263928e-01 -6.84965193e-01 1.12145519e+00
1.62439391e-01 -1.25781512e+00 -1.17189191e-01 9.30754170e-02
4.39215928e-01 -2.22765565e-01 1.33982778e-01 -4.49185818e-02
-2.12617785e-01 -7.37950563e-01 8.68633807e-01 6.05184674e-01
7.85594523e-01 -3.47618639e-01 8.06158304e-01 5.65966249e-01
-1.50777650e+00 -1.83600277e-01 -1.95331901e-01 -4.30429913e-02
1.45059243e-01 6.70820534e-01 -6.78219497e-01 3.19337904e-01
8.81835222e-01 5.66955030e-01 -5.76573610e-01 1.28745925e+00
-2.40901157e-01 4.69815254e-01 -3.86993945e-01 7.18537644e-02
6.32754713e-02 -2.45516911e-01 8.96297514e-01 9.39425349e-01
1.81944683e-01 2.92169005e-01 3.97788227e-01 1.09554994e+00
2.02437058e-01 4.66746762e-02 -2.39064440e-01 3.16013128e-01
5.17752171e-01 1.51186705e+00 -9.76505458e-01 -5.38765252e-01
1.53878973e-05 1.12714040e+00 -1.65690761e-03 4.78203386e-01
-1.09592569e+00 -5.20984411e-01 5.19844890e-01 2.43118420e-01
6.86156869e-01 1.04708470e-01 -2.20429346e-01 -1.15147686e+00
1.36583745e-01 -5.95531642e-01 4.23250377e-01 -1.14533484e+00
-9.78923798e-01 3.85091573e-01 -2.13186264e-01 -1.29417646e+00
6.95228755e-01 -3.65258157e-01 -8.86117637e-01 8.65316868e-01
-1.36467707e+00 -1.19647610e+00 -7.70153403e-01 4.38761950e-01
5.23432553e-01 2.69922256e-01 2.53900260e-01 3.19113225e-01
-7.46824324e-01 2.86187351e-01 -3.14276963e-01 2.03508571e-01
5.93402088e-01 -6.10811889e-01 -2.69153249e-02 1.02537966e+00
-3.05442065e-01 5.00547111e-01 6.89654946e-01 -1.09724951e+00
-1.29943466e+00 -1.12028837e+00 5.57721972e-01 -4.77092475e-01
4.17370081e-01 -3.99304509e-01 -1.24521720e+00 3.12958658e-01
-6.17993064e-02 2.97768831e-01 -1.81536719e-01 -5.02043307e-01
-1.21962465e-02 -4.57589068e-02 -1.16758907e+00 7.20642745e-01
1.12909818e+00 -1.42198369e-01 -6.32571518e-01 3.91505688e-01
1.02510571e+00 -4.57584381e-01 -5.22728898e-02 4.61305916e-01
4.04132128e-01 -8.24936867e-01 8.53227317e-01 -1.09983124e-01
-5.55561893e-02 -8.92840564e-01 1.10739477e-01 -6.37101710e-01
-3.72945845e-01 -5.97596765e-01 -2.70008296e-01 1.27067769e+00
-1.48137808e-01 -6.37934029e-01 7.98583746e-01 3.58257055e-01
-1.17035173e-01 -8.50262344e-01 -1.12755096e+00 -4.83560055e-01
-6.36068761e-01 -8.89186934e-02 6.75624669e-01 8.41860771e-01
-5.15345633e-01 2.10709721e-01 -4.05427009e-01 3.51797819e-01
6.00125074e-01 5.21253884e-01 7.29694188e-01 -1.19860661e+00
2.79203445e-01 -4.62477714e-01 -1.50695533e-01 -1.44421208e+00
-7.32481331e-02 -5.58760464e-01 1.46213546e-01 -1.42519927e+00
2.89912939e-01 -5.14829695e-01 -2.74194211e-01 3.93953770e-01
-4.57770288e-01 2.88442999e-01 4.44746651e-02 4.02110010e-01
-8.39294374e-01 6.28018260e-01 1.77532971e+00 -2.92246968e-01
-3.25360805e-01 -1.00048259e-02 -5.71627140e-01 8.24502289e-01
6.30214095e-01 -5.66573918e-01 -7.16912374e-02 -3.19512665e-01
-2.97045559e-01 -6.16437048e-02 8.52055073e-01 -6.57701135e-01
4.75060523e-01 -3.53842884e-01 5.96225262e-01 -1.01770353e+00
5.29756427e-01 -8.04362357e-01 -1.10296011e-01 7.09388912e-01
2.14649573e-01 -2.64121830e-01 1.65976241e-01 8.24517965e-01
1.41175121e-01 -6.40798211e-02 7.65491903e-01 -1.77329972e-01
-8.77298653e-01 4.86081660e-01 -2.69102514e-01 3.22506428e-02
1.21857774e+00 -5.55368721e-01 -6.13645375e-01 -5.32551035e-02
-4.94387954e-01 4.09715116e-01 5.26844025e-01 4.19351339e-01
8.29511166e-01 -1.10979021e+00 -6.02025568e-01 4.91078794e-01
-4.51659262e-02 2.19683930e-01 4.42414939e-01 1.22911990e+00
-1.20581947e-01 1.04282729e-01 1.36735022e-01 -8.03367198e-01
-1.57822871e+00 1.00620592e+00 5.59308469e-01 2.52235144e-01
-7.20733345e-01 1.10569656e+00 8.24657321e-01 2.65486747e-01
2.16181725e-01 -3.76412064e-01 -1.95133433e-01 6.71118051e-02
9.05255079e-01 4.27694947e-01 -3.67186338e-01 -7.70487964e-01
-6.82617605e-01 8.20564926e-01 -8.14438984e-02 2.77635545e-01
6.55262887e-01 -5.01871049e-01 -4.75580752e-01 1.76458359e-01
1.11629450e+00 1.85732558e-01 -1.25171053e+00 -4.13297206e-01
-2.28669941e-01 -8.91043425e-01 2.31701881e-01 -7.79451966e-01
-1.21432459e+00 7.76761174e-01 7.63612628e-01 7.49193430e-02
1.27689505e+00 2.56710172e-01 1.02376604e+00 5.93938231e-02
4.13053393e-01 -7.25903034e-01 1.61088035e-01 2.21812651e-01
8.35834444e-01 -1.18035710e+00 1.12553127e-02 -6.58151925e-01
-4.63389993e-01 7.21140087e-01 8.61229539e-01 -6.84487671e-02
6.50024295e-01 2.30356082e-01 7.81265423e-02 -3.61100525e-01
-5.42490184e-01 -4.81924236e-01 4.40291375e-01 4.47558969e-01
-4.41664636e-01 -1.24287613e-01 -1.19429402e-01 7.49815822e-01
2.74215251e-01 -1.03427023e-01 -5.68920979e-04 7.42311358e-01
-9.36105907e-01 -4.44827884e-01 -7.66081214e-01 3.06757361e-01
-4.01570052e-01 -9.16528106e-02 -3.98978978e-01 5.51830113e-01
7.62745798e-01 1.15073740e+00 2.28960305e-01 -3.62916708e-01
-5.73414937e-02 -3.21713775e-01 2.04239801e-01 -4.02301192e-01
-3.09010923e-01 4.61217016e-01 -2.86761045e-01 -5.12400806e-01
-1.12544172e-01 -4.77896124e-01 -1.41184855e+00 -2.31565889e-02
-9.28027630e-01 3.11801940e-01 2.55445570e-01 1.01540828e+00
3.82303089e-01 3.29085737e-01 4.92728770e-01 -9.99416411e-01
2.47573201e-02 -8.41447055e-01 -4.61865038e-01 2.99516201e-01
6.50423408e-01 -9.94305670e-01 -5.79604030e-01 1.28786936e-01] | [9.507146835327148, -0.2824445068836212] |
351f6760-b21c-447f-adaf-515849549c6f | multi-scale-attention-guided-pose-transfer | 2202.06777 | null | https://arxiv.org/abs/2202.06777v1 | https://arxiv.org/pdf/2202.06777v1.pdf | Multi-scale Attention Guided Pose Transfer | Pose transfer refers to the probabilistic image generation of a person with a previously unseen novel pose from another image of that person having a different pose. Due to potential academic and commercial applications, this problem is extensively studied in recent years. Among the various approaches to the problem, attention guided progressive generation is shown to produce state-of-the-art results in most cases. In this paper, we present an improved network architecture for pose transfer by introducing attention links at every resolution level of the encoder and decoder. By utilizing such dense multi-scale attention guided approach, we are able to achieve significant improvement over the existing methods both visually and analytically. We conclude our findings with extensive qualitative and quantitative comparisons against several existing methods on the DeepFashion dataset. | ['Umapada Pal', 'Subhankar Ghosh', 'Saumik Bhattacharya', 'Prasun Roy'] | 2022-02-14 | null | null | null | null | ['pose-transfer'] | ['computer-vision'] | [ 4.85456318e-01 4.16638136e-01 2.57450670e-01 -2.06332833e-01
-9.32101011e-01 -2.51994759e-01 7.28511155e-01 -2.67381221e-01
-4.03068691e-01 9.43786323e-01 5.07155776e-01 4.99232650e-01
2.36954242e-02 -5.93476236e-01 -8.93416584e-01 -3.67791295e-01
-2.83069797e-02 6.70887291e-01 -1.65531542e-02 -2.11965114e-01
2.99518667e-02 4.17616099e-01 -1.61346281e+00 2.20477074e-01
4.80512917e-01 9.94353831e-01 1.31692559e-01 6.20735586e-01
4.63988543e-01 4.43252057e-01 -8.11450541e-01 -8.59184682e-01
2.95246154e-01 -5.09250522e-01 -9.32030916e-01 1.88741565e-01
1.13130629e+00 -5.03831208e-01 -5.51361263e-01 8.36890996e-01
9.01124477e-01 1.55836910e-01 6.01437628e-01 -1.14281428e+00
-1.04819798e+00 3.91502529e-01 -7.30185747e-01 1.93283483e-01
6.76322162e-01 8.95270035e-02 8.76947105e-01 -1.01771319e+00
8.84606421e-01 1.60225153e+00 4.85768020e-01 7.12141573e-01
-1.34325707e+00 -5.39965987e-01 2.99376070e-01 3.78044069e-01
-1.55769062e+00 -2.11686283e-01 7.10624099e-01 -3.63935709e-01
8.48935366e-01 1.75800510e-02 7.10168839e-01 1.60289407e+00
2.50992149e-01 8.40246260e-01 1.07497287e+00 -2.69257605e-01
-1.92231327e-01 -5.74675500e-02 -4.17519212e-01 6.21392310e-01
1.81122258e-01 3.62363219e-01 -7.87985742e-01 2.07936317e-01
8.90165806e-01 -2.19993070e-01 -2.19592541e-01 -3.81253392e-01
-1.19268477e+00 6.85135186e-01 9.67189193e-01 -2.35256925e-02
-4.34370428e-01 5.69131434e-01 1.16069734e-01 -2.09767193e-01
5.98818779e-01 6.08086050e-01 -5.77840172e-02 -3.19547765e-02
-1.03884995e+00 6.99311018e-01 3.58893037e-01 9.86518085e-01
3.03069621e-01 5.18298596e-02 -6.90797031e-01 4.75781500e-01
3.22401851e-01 4.52197641e-01 1.08147398e-01 -9.35413241e-01
3.91736239e-01 3.34044099e-01 2.62735456e-01 -9.91076410e-01
-2.68204212e-01 -8.30867529e-01 -7.63475120e-01 2.58387476e-01
1.94307148e-01 -3.31580907e-01 -1.10175395e+00 1.89793277e+00
2.36805722e-01 2.32706666e-01 -1.84083983e-01 1.06418192e+00
8.33069742e-01 7.15876102e-01 1.39793590e-01 1.73449397e-01
1.37578428e+00 -1.03929114e+00 -7.68874347e-01 -4.37937498e-01
-3.13934237e-01 -7.15802491e-01 8.36895585e-01 3.04364622e-01
-1.24941695e+00 -8.13123524e-01 -9.35847223e-01 -3.55606467e-01
-1.42121598e-01 2.06350610e-01 5.65398633e-01 4.17799085e-01
-1.31450152e+00 5.01952589e-01 -6.39243126e-01 -5.82469165e-01
6.62681103e-01 3.43439013e-01 -3.23900312e-01 1.68927722e-02
-1.18157172e+00 9.69496489e-01 1.50418505e-01 1.86613962e-01
-9.78426397e-01 -5.53974032e-01 -8.73210490e-01 7.29303584e-02
2.71185935e-01 -1.24223292e+00 1.33055854e+00 -7.73262143e-01
-1.29161870e+00 7.88812935e-01 -1.78889766e-01 -7.27452517e-01
7.46596158e-01 -7.19049037e-01 -1.87276110e-01 2.83949405e-01
1.96505100e-01 1.46248567e+00 8.20347011e-01 -1.31828344e+00
-5.08354366e-01 -3.34534585e-01 2.24838957e-01 4.54840422e-01
-1.99223563e-01 7.25927874e-02 -5.93812823e-01 -1.11391866e+00
-1.89101622e-01 -1.05196190e+00 -1.77204445e-01 2.09845096e-01
-6.77916706e-01 -2.27083907e-01 6.71733677e-01 -8.12956274e-01
8.40946615e-01 -1.78641593e+00 7.10286915e-01 -2.41768479e-01
2.35271305e-01 1.15462154e-01 -3.10142133e-02 4.36509937e-01
-7.51407295e-02 9.09594372e-02 -5.57230264e-02 -7.65851378e-01
-2.23418530e-02 -2.40522206e-01 -2.27355108e-01 3.67825508e-01
4.27091897e-01 1.28006780e+00 -7.05297589e-01 -4.55257863e-01
2.60277152e-01 1.00624049e+00 -6.46782219e-01 2.38133729e-01
-4.74258550e-02 6.78781688e-01 -1.34990498e-01 4.08197761e-01
3.60278070e-01 -3.10749561e-01 -2.43016124e-01 -3.03189784e-01
2.00108364e-01 -2.24329326e-02 -8.90786231e-01 1.89813435e+00
-2.92266458e-01 8.57564807e-01 -2.09598020e-01 -4.89752978e-01
5.66230834e-01 3.13140094e-01 2.35475942e-01 -5.24643362e-01
2.90848732e-01 -1.92746133e-01 5.60863689e-03 -2.00201690e-01
8.49548936e-01 -3.25466543e-01 -1.05442822e-01 6.23832382e-02
2.90487587e-01 2.18784705e-01 2.06551597e-01 2.12175816e-01
7.59193242e-01 4.41866636e-01 1.21574827e-01 -1.23040497e-01
3.57984394e-01 -2.80178487e-01 1.06060050e-01 6.36580408e-01
-1.89797923e-01 1.16771603e+00 1.44269615e-01 -3.51268858e-01
-1.12213755e+00 -1.32962978e+00 -1.00523205e-02 9.50895309e-01
2.17282921e-01 -4.17727590e-01 -8.95524859e-01 -5.31739295e-01
-1.75671339e-01 4.92643416e-01 -9.11201179e-01 -1.01850249e-01
-6.08042121e-01 -3.50225002e-01 4.00487125e-01 9.41242337e-01
8.15335035e-01 -1.31086636e+00 -4.55954462e-01 1.23047076e-01
-4.07841384e-01 -1.17948306e+00 -4.53274965e-01 -4.65082169e-01
-4.66474682e-01 -8.01702321e-01 -1.29098201e+00 -8.52589548e-01
6.64944887e-01 -5.26534102e-04 1.16645849e+00 -2.08193988e-01
-5.82719386e-01 3.82625222e-01 -2.37755075e-01 -3.57789516e-01
-7.49976858e-02 2.62341291e-01 4.54399437e-02 1.20406903e-01
1.84155647e-02 -3.74833345e-01 -9.18328762e-01 2.34077796e-02
-7.79569745e-01 2.84032136e-01 7.89483726e-01 6.81454360e-01
5.20253181e-01 -1.70612887e-01 5.68785191e-01 -5.80092788e-01
5.67887366e-01 -1.83803856e-01 -3.16329807e-01 -1.77642312e-02
-2.51704931e-01 4.52974066e-02 2.59578824e-01 -1.68010890e-01
-1.08122993e+00 2.64688134e-01 -2.95464277e-01 -3.70689124e-01
-3.21354449e-01 1.39307126e-01 -1.29275456e-01 4.13679555e-02
6.80210233e-01 1.77402243e-01 -1.17579885e-01 -3.89914751e-01
5.58950424e-01 3.25585932e-01 7.06950724e-01 -3.28435481e-01
1.06894410e+00 5.96886456e-01 -4.24519666e-02 -6.35128736e-01
-1.04702187e+00 -7.51349702e-02 -7.35080481e-01 -4.30363506e-01
1.29870951e+00 -1.02707982e+00 -6.68612838e-01 3.79044741e-01
-1.33239245e+00 -6.42239302e-02 -2.35180974e-01 2.27803379e-01
-6.22063637e-01 1.06039509e-01 -5.71878850e-01 -6.79067492e-01
-2.85049677e-01 -1.14227962e+00 1.54460049e+00 3.34273279e-01
-4.84008431e-01 -8.00287127e-01 5.02558313e-02 4.67527092e-01
3.49051505e-01 5.22025943e-01 4.01103109e-01 -1.95773035e-01
-8.06444407e-01 -3.14333409e-01 -2.82675713e-01 1.89096496e-01
-7.50898421e-02 -4.13167357e-01 -9.38140690e-01 -6.02068126e-01
-4.71556157e-01 -3.34911466e-01 9.60850894e-01 4.11746234e-01
1.10230553e+00 -2.33870953e-01 -4.50631529e-01 4.48822021e-01
1.19895637e+00 -1.56595707e-01 8.51598561e-01 1.88821465e-01
8.67790222e-01 5.84704638e-01 4.41540718e-01 2.83087879e-01
3.04760486e-01 9.74391937e-01 4.21539515e-01 -3.47367197e-01
-6.15182877e-01 -4.38770682e-01 1.66953266e-01 -1.08914450e-01
-4.86944914e-01 -5.99035203e-01 -6.77857757e-01 5.23599505e-01
-1.77380788e+00 -1.13495839e+00 2.34506577e-01 2.05447745e+00
5.45833707e-01 9.10100266e-02 1.99320599e-01 -5.80094606e-02
9.56328928e-01 1.73007131e-01 -4.20093000e-01 -4.31680307e-03
-4.26323116e-02 3.66246641e-01 2.44577214e-01 3.85529578e-01
-1.27957380e+00 9.41024959e-01 7.36612558e+00 6.41040802e-01
-7.47069180e-01 -8.58675539e-02 7.60939121e-01 -3.58700395e-01
-4.06600833e-02 -3.86345267e-01 -9.61944759e-01 2.67447084e-01
6.87641859e-01 -1.78844973e-01 2.51233310e-01 6.75388753e-01
6.31338134e-02 -1.47435935e-02 -1.37845409e+00 9.86008883e-01
5.40160060e-01 -1.21885216e+00 3.76071781e-01 2.75869578e-01
9.04968858e-01 -3.93706679e-01 6.22411251e-01 1.83547750e-01
8.71452838e-02 -1.33827853e+00 9.93477643e-01 6.29961491e-01
9.60703611e-01 -9.24450040e-01 7.32913911e-01 -1.10618286e-02
-1.11851382e+00 2.08574384e-02 -7.30268806e-02 -1.52299434e-01
5.76022506e-01 2.78883219e-01 -6.64828062e-01 5.47919929e-01
8.06942284e-01 6.13194764e-01 -8.12174618e-01 1.14489567e+00
-4.90955144e-01 2.34467655e-01 -4.23273332e-02 2.05832198e-01
2.38887053e-02 2.90400714e-01 7.43983269e-01 9.97530580e-01
4.30030674e-01 -2.35768393e-01 4.08068225e-02 1.04768574e+00
-2.21525088e-01 -2.11167037e-01 -6.22368753e-01 2.82444149e-01
1.45409316e-01 1.10835040e+00 -5.87797046e-01 -4.32994843e-01
-1.40138432e-01 1.40912533e+00 4.21092480e-01 3.02566737e-01
-1.14045191e+00 -2.42976636e-01 5.91861725e-01 5.47750294e-01
5.18686116e-01 -2.58677691e-01 1.70209948e-02 -7.41731465e-01
-6.91921916e-03 -5.50275922e-01 1.69823423e-01 -9.82367992e-01
-1.32661366e+00 9.31993365e-01 2.81579226e-01 -1.11898625e+00
-5.14576733e-01 -5.32825351e-01 -2.54853845e-01 9.40964758e-01
-1.13634849e+00 -1.42171776e+00 -3.85022372e-01 3.18657637e-01
7.79278219e-01 -1.81600198e-01 7.54544914e-01 2.45431527e-01
-3.16143751e-01 7.61929691e-01 -4.28546697e-01 2.74135433e-02
7.76236594e-01 -1.29604924e+00 7.74045229e-01 1.05204380e+00
2.54031897e-01 6.11187518e-01 8.53992879e-01 -7.81910896e-01
-8.49072635e-01 -1.30556738e+00 9.04809773e-01 -5.71028531e-01
2.04673037e-01 -5.35415471e-01 -4.44272906e-01 8.03256929e-01
6.23037815e-01 -1.91308066e-01 4.30902332e-01 -2.85749622e-02
-1.76857591e-01 1.05525643e-01 -1.04427433e+00 9.18560147e-01
1.31700313e+00 -2.67558157e-01 -6.04131222e-01 3.07246596e-01
7.19713390e-01 -6.17420316e-01 -5.96619368e-01 3.77088308e-01
5.04224777e-01 -1.12028432e+00 1.18659449e+00 -4.07533646e-01
9.10670817e-01 -2.27508724e-01 9.98025537e-02 -1.37079680e+00
-7.75506258e-01 -5.90397835e-01 -3.19422215e-01 1.07337606e+00
2.86792666e-01 -2.53517509e-01 8.63331437e-01 3.85960370e-01
-2.05708161e-01 -7.94375122e-01 -8.97408307e-01 -7.03250527e-01
5.98195530e-02 -1.21480517e-01 3.39380920e-01 1.58310115e-01
-3.98030341e-01 6.89470291e-01 -7.44694293e-01 1.84961677e-01
7.12197125e-01 -1.02724113e-01 7.56102085e-01 -1.05764377e+00
-3.36106449e-01 -2.95000762e-01 -7.27907896e-01 -1.22044909e+00
1.48714767e-04 -7.93191135e-01 1.04460977e-01 -2.01396632e+00
4.04990941e-01 2.67428428e-01 -8.80149528e-02 2.72935182e-01
-2.99119085e-01 9.26212609e-01 4.09011334e-01 -1.39323995e-01
-6.72010720e-01 7.55080163e-01 1.45100641e+00 -1.24874465e-01
1.33007377e-01 -1.24193616e-01 -1.00681281e+00 6.42680347e-01
6.26692712e-01 -2.07309484e-01 -4.44251418e-01 -4.99183446e-01
1.82433158e-01 -2.26201028e-01 7.51909077e-01 -1.45487070e+00
-7.82720447e-02 2.70741373e-01 9.68653858e-01 -6.82029307e-01
9.23668087e-01 -5.78893960e-01 1.49587914e-01 5.06902874e-01
-4.38831478e-01 1.79502413e-01 1.86721131e-01 8.15387249e-01
-8.39989558e-02 3.03041071e-01 8.43439281e-01 -7.25921690e-02
-6.71461225e-01 4.57295358e-01 -1.66945681e-01 4.13186513e-02
1.13517785e+00 -1.44066781e-01 -1.88764662e-01 -7.02526510e-01
-9.86708581e-01 5.33197410e-02 2.74245650e-01 7.06119537e-01
8.18115711e-01 -1.63673365e+00 -9.82015491e-01 5.54635823e-02
1.71390355e-01 -6.38481155e-02 3.84249091e-01 5.22542655e-01
-3.75315666e-01 5.20010293e-01 -3.36228400e-01 -4.97543931e-01
-1.17339623e+00 4.80302811e-01 2.67200261e-01 -2.04866782e-01
-6.33830428e-01 1.09374583e+00 5.99858046e-01 1.10275470e-01
2.83574134e-01 -3.18163447e-02 -1.82374835e-01 -1.70963466e-01
5.71570456e-01 3.69359404e-01 -2.33042717e-01 -1.15172458e+00
-3.37363064e-01 5.98893523e-01 -2.11622044e-01 -3.95267934e-01
1.06651831e+00 -1.68756157e-01 3.98409724e-01 2.11133480e-01
1.04683244e+00 -1.73289925e-01 -1.64290810e+00 4.88795340e-02
-6.66641712e-01 -4.81938422e-01 -2.63963312e-01 -1.11562800e+00
-1.06041110e+00 8.18232596e-01 7.42034078e-01 -1.55748799e-01
9.37822938e-01 3.17046672e-01 7.85836995e-01 7.08373338e-02
4.84645486e-01 -9.02185261e-01 5.32932103e-01 3.25857222e-01
1.39989436e+00 -1.16931546e+00 5.19641908e-03 -5.03533304e-01
-7.02614665e-01 5.63302457e-01 8.97045910e-01 -4.15079355e-01
3.16031843e-01 -1.08799133e-02 -2.02798173e-01 -1.57957956e-01
-4.66881245e-01 -2.65771955e-01 7.95605123e-01 8.83063078e-01
6.00413680e-01 -7.34200850e-02 -1.60410985e-01 4.72114176e-01
-6.37571037e-01 -2.27757454e-01 2.88959801e-01 6.26927018e-01
-3.30603361e-01 -9.39510345e-01 -4.39922839e-01 3.20705295e-01
-4.51184630e-01 -9.73321870e-02 -6.52799785e-01 8.14886034e-01
1.93224207e-01 7.37816334e-01 1.27824187e-01 -3.30903590e-01
4.53341097e-01 -1.18417285e-01 9.18503821e-01 -7.30142236e-01
-4.30777907e-01 1.07910998e-01 6.41616136e-02 -6.25654161e-01
-3.70062709e-01 -6.69103205e-01 -1.05305445e+00 -1.09064251e-01
1.15365796e-02 -3.85211259e-01 3.28548819e-01 7.62293398e-01
6.19448841e-01 9.77320611e-01 3.16349387e-01 -1.43704140e+00
-1.47561938e-01 -1.10364628e+00 -3.36052090e-01 5.71397305e-01
1.27601475e-01 -1.09013987e+00 2.31885258e-02 2.43648753e-01] | [11.906764030456543, -0.7405266761779785] |
76747dc6-e537-4a79-b1ab-6d8fb16aaae1 | wppg-net-a-non-contact-video-based-heart-rate | 2207.01697 | null | https://arxiv.org/abs/2207.01697v2 | https://arxiv.org/pdf/2207.01697v2.pdf | BYHE: A Simple Framework for Boosting End-to-end Video-based Heart Rate Measurement Network | Heart rate measuring based on remote photoplethysmography (rPPG) plays an important role in health caring, which estimates heart rate from facial video in a non-contact, less-constrained way. End-to-end neural network is a main branch of rPPG-based heart rate estimation methods, whose trait is recovering rPPG signal containing sufficient heart rate message from original facial video directly. However, there exists some easily neglected problems on relevant datasets which thwarting the efficient training of end-to-end methods, such as uncertain temporal delay and indefinite envelope shape of label waves. Although many novel and powerful networks are proposed, hitherto there are no systematic research digging into these problems. In this paper, from perspective of common intrinsic rhythm periodical self-similarity results from cardiac activities, we propose a comprehensive methodology, Boost Your Heartbeat Estimation (BYHE), including new label representations, corresponding network adjustments and loss functions. BYHE can be easily grafted on current end-to-end network and boost its training efficiency. By applying our methodology, we can save tremendous time without conducting laborious handworks, such as label wave alignment which is necessary for previous end-to-end methods, and meanwhile enhance the utilization on datasets. According to our experiments, BYHE can leverage classical end-to-end network to reach competitive performance against those state-of-the-art methods on mostly used datasets. Such improvement indicates selecting perspicuous and efficient label representation is also a promising direction towards better remote physiological signal measurement. | ['Xiaolin Huang', 'Chunyu Ji', 'Yun Ge', 'Ying Chen', 'Xinyu Zhang', 'Weiyu Sun'] | 2022-07-04 | null | null | null | null | ['heart-rate-estimation'] | ['medical'] | [ 1.55840367e-01 5.49206464e-03 -3.03820461e-01 -4.01109844e-01
-2.80763239e-01 -1.11492388e-01 -1.70792654e-01 -6.35098100e-01
-2.29855314e-01 8.18348944e-01 5.31419972e-03 2.55774852e-04
-1.21002316e-01 -3.60704660e-01 1.05520887e-02 -8.98186386e-01
-1.29504308e-01 -1.70326829e-01 -3.11322927e-01 -1.97481677e-01
-6.33899346e-02 3.92900199e-01 -1.29129684e+00 -2.68135190e-01
5.90242267e-01 1.18164337e+00 -3.42681170e-01 5.65277278e-01
1.90349132e-01 7.42405951e-01 -3.39285403e-01 -1.94178954e-01
1.40654996e-01 -8.99255514e-01 -2.96673924e-01 -3.94900978e-01
1.73331648e-01 -3.97539228e-01 -3.42011690e-01 7.23030090e-01
1.35585535e+00 -2.60198891e-01 4.42717433e-01 -1.16214216e+00
-2.79681772e-01 3.59754652e-01 -7.28024185e-01 3.54341775e-01
2.74205524e-02 2.82542735e-01 4.49486196e-01 -5.09129345e-01
4.23302263e-01 5.84413111e-01 1.22697055e+00 8.46100152e-01
-8.19271803e-01 -1.04616046e+00 -3.59321654e-01 2.91192710e-01
-1.45711112e+00 -7.04624474e-01 1.29337847e+00 -8.22400749e-02
3.39614123e-01 3.64409357e-01 9.10123467e-01 1.02338374e+00
-4.95693237e-02 3.87640685e-01 1.16967630e+00 -1.44562364e-01
-1.82273343e-01 1.34758338e-01 -2.13902786e-01 8.09034169e-01
-1.94542542e-01 2.98951924e-01 -4.61826324e-01 1.07825205e-01
9.70220149e-01 -1.02696970e-01 -7.99946666e-01 -1.17787570e-02
-1.27405310e+00 4.06307906e-01 1.25039011e-01 3.84453595e-01
-2.75721043e-01 1.57318443e-01 7.60782003e-01 2.82348990e-01
5.76213896e-01 9.80960578e-02 -5.30297160e-01 -5.41359425e-01
-1.05245638e+00 -3.05800259e-01 7.22864807e-01 5.80756843e-01
4.87036943e-01 3.15902263e-01 -3.17743689e-01 7.94714630e-01
1.45667300e-01 5.99436104e-01 6.53718650e-01 -9.70146239e-01
3.79496887e-02 2.47760221e-01 -2.16054842e-01 -1.27744699e+00
-8.71284783e-01 -6.70003295e-01 -1.35574996e+00 8.66194144e-02
3.52890253e-01 -5.11370540e-01 -5.14122784e-01 1.91570830e+00
5.12896180e-01 7.92065024e-01 -1.25228360e-01 1.37148952e+00
1.14899790e+00 5.19626975e-01 1.00197412e-01 -1.06592071e+00
1.33025944e+00 -6.86963618e-01 -9.72301006e-01 2.24078014e-01
5.89883149e-01 -5.69341958e-01 7.80637741e-01 5.19755006e-01
-9.16814744e-01 -7.57303715e-01 -8.44330430e-01 7.86145255e-02
1.40257016e-01 3.03508133e-01 6.69437289e-01 9.66454029e-01
-1.03899956e+00 9.23133194e-01 -3.96028131e-01 -2.42735535e-01
3.95772070e-01 2.66856283e-01 -3.37287188e-01 2.03393489e-01
-1.52739954e+00 6.49600267e-01 9.29949582e-02 8.18759918e-01
-5.71664572e-01 -8.26615393e-01 -5.60095727e-01 -2.90934861e-01
3.59238654e-01 -7.15351582e-01 8.88850272e-01 -8.92649770e-01
-2.04467463e+00 9.05484617e-01 1.73424501e-02 -1.44109234e-01
7.15293765e-01 1.29395455e-01 -6.66291893e-01 3.62492859e-01
-5.01024425e-01 4.24263626e-01 1.08817828e+00 -7.64575362e-01
-8.82262588e-02 -3.32270056e-01 -5.74261844e-01 4.33857083e-01
-3.97466570e-01 -7.22058490e-03 -2.95363396e-01 -3.47539425e-01
1.30259320e-01 -6.95180893e-01 1.49644524e-01 3.69280398e-01
-5.48111908e-02 -1.37973472e-01 7.29670167e-01 -9.17031944e-01
1.33606052e+00 -2.08065891e+00 -1.33803904e-01 5.64380921e-02
5.61446786e-01 6.35100961e-01 -5.13137989e-02 5.45209795e-02
-3.18102270e-01 1.69009611e-01 4.22166213e-02 -1.78266436e-01
-2.30497569e-01 -1.93617523e-01 3.84481549e-02 8.90799463e-01
-2.40170792e-01 8.40112507e-01 -6.30271018e-01 -7.50481606e-01
3.03615153e-01 6.93447828e-01 -7.93017447e-02 3.23261619e-01
4.91114289e-01 7.00227082e-01 -2.08874986e-01 6.74859703e-01
9.04381871e-01 -4.65187915e-02 1.05622392e-02 -7.49480247e-01
-2.93281246e-02 -3.80390584e-01 -1.00362802e+00 1.72917855e+00
-5.02639234e-01 6.22478426e-01 -2.59498619e-02 -1.20253992e+00
1.41393697e+00 6.77616417e-01 7.91586578e-01 -7.39781141e-01
4.59506422e-01 1.01326875e-01 -6.11669235e-02 -1.16421819e+00
-2.24002808e-01 -5.54454327e-01 3.36779952e-01 3.60968679e-01
-8.61947909e-02 2.94680417e-01 -1.67596266e-01 -2.44294286e-01
6.97920978e-01 3.48937690e-01 2.91587979e-01 -1.56169534e-02
6.88977182e-01 -6.75080895e-01 8.65843236e-01 3.89668137e-01
-8.09075296e-01 8.86552811e-01 5.63811362e-01 -6.89670324e-01
-8.39385092e-01 -5.06341338e-01 -2.52863646e-01 6.30765855e-01
3.65471654e-02 -1.41458511e-01 -5.87016404e-01 -5.44234097e-01
-3.16171974e-01 -9.91550982e-02 -6.14439368e-01 -2.04819798e-01
-6.33403957e-01 -9.15617406e-01 1.08759880e+00 4.19851333e-01
7.61966527e-01 -1.21868145e+00 -6.66506708e-01 2.49434888e-01
-4.52566892e-01 -9.80581999e-01 -4.58520025e-01 -2.68371254e-01
-1.01176775e+00 -9.00817275e-01 -1.22052550e+00 -5.29018641e-01
2.62860596e-01 1.65218428e-01 1.04589200e+00 1.62577301e-01
-6.55682981e-01 1.40578985e-01 -2.93832481e-01 -3.55594307e-01
-4.26658317e-02 1.92980953e-02 4.63684537e-02 2.76552618e-01
2.34123185e-01 -9.56715465e-01 -1.18375993e+00 3.76973003e-01
-2.97244757e-01 7.98929110e-02 5.50240993e-01 5.27374864e-01
3.16993952e-01 -2.24238142e-01 1.06544518e+00 -6.25925660e-01
6.01654172e-01 -2.59718865e-01 -3.67497236e-01 1.45488739e-01
-9.39537108e-01 -4.62482810e-01 7.26867199e-01 -5.32723844e-01
-9.31701243e-01 -1.64186642e-01 -2.72835284e-01 -6.63451493e-01
-6.68786094e-02 4.72829491e-01 1.75175369e-01 -2.27407843e-01
6.81687355e-01 4.61223155e-01 6.38934195e-01 -6.10688366e-02
5.30233458e-02 7.21604109e-01 6.66094244e-01 -2.20351234e-01
5.05053043e-01 3.71858060e-01 3.45720679e-01 -7.62585998e-01
-5.20212352e-01 -5.35691679e-01 -3.28341663e-01 -6.37991488e-01
7.31713593e-01 -1.01718807e+00 -1.38962328e+00 7.43704617e-01
-9.82160687e-01 -1.06319427e-01 -1.26595404e-02 6.47740304e-01
-4.93847162e-01 6.38246119e-01 -6.06269538e-01 -1.06678855e+00
-1.02436769e+00 -4.89448220e-01 5.59328318e-01 6.31730258e-01
7.90239722e-02 -1.00641215e+00 2.99312919e-01 4.04441565e-01
8.19388568e-01 5.77656507e-01 4.19947475e-01 5.25792455e-03
-1.37538433e-01 -2.43213937e-01 -4.09848481e-01 5.73795557e-01
1.25429071e-02 5.15009873e-02 -1.33237684e+00 -2.24625856e-01
3.05885732e-01 -3.61159980e-01 6.48166776e-01 6.41130924e-01
1.23862410e+00 -8.36278424e-02 -1.09381318e-01 1.06353784e+00
1.42747653e+00 1.37898684e-01 1.13355958e+00 -1.32670149e-01
7.67322361e-01 6.45944655e-01 4.46222931e-01 6.13479197e-01
2.14927852e-01 3.67753446e-01 3.27377677e-01 -6.99921608e-01
-2.14991435e-01 1.54606551e-02 2.87051320e-01 9.97435808e-01
-5.79442561e-01 1.69675022e-01 -4.79812145e-01 2.33507022e-01
-1.67820108e+00 -1.13176787e+00 -8.91710892e-02 2.27690339e+00
1.03872538e+00 -2.96952635e-01 3.11138570e-01 2.36544058e-01
9.00619268e-01 4.15614277e-01 -7.70947218e-01 -1.93738237e-01
7.99730122e-02 2.12955877e-01 2.63719350e-01 -8.47142190e-03
-9.14290249e-01 3.34073722e-01 5.46422005e+00 7.86858857e-01
-1.69317877e+00 2.58700460e-01 8.21570694e-01 -1.36781663e-01
2.67707288e-01 -2.29295835e-01 -7.31283307e-01 5.98946512e-01
1.00954223e+00 -1.46998510e-01 4.85706329e-01 6.61112249e-01
7.04678893e-01 1.62298605e-01 -7.82423019e-01 1.66712010e+00
3.56764019e-01 -1.18207824e+00 -7.73197055e-01 -2.55747110e-01
9.74625573e-02 -3.52346450e-01 -1.93230122e-01 3.76575202e-01
-9.17997777e-01 -9.41798568e-01 -4.39310111e-02 7.65217781e-01
1.42498410e+00 -7.41223812e-01 9.87877727e-01 2.46190652e-01
-1.11516082e+00 1.03287771e-01 -4.35831487e-01 3.72364372e-02
5.07480651e-02 8.05526018e-01 -6.15940690e-01 6.19622350e-01
3.53759080e-01 9.83325541e-01 -1.68537110e-01 1.12639797e+00
-6.57868981e-02 6.64550006e-01 -2.01567039e-01 -8.24022442e-02
-2.95830786e-01 -2.36679792e-01 2.19707727e-01 9.88561869e-01
3.79413068e-01 3.63573313e-01 -1.52990952e-01 7.46427953e-01
-5.21645248e-02 2.53544927e-01 -6.20813131e-01 2.22992048e-01
2.61387855e-01 1.75331140e+00 -5.44475377e-01 -1.25698656e-01
-2.12957159e-01 6.55636609e-01 -1.18761681e-01 2.68448412e-01
-1.18598998e+00 -7.87411630e-01 2.21012279e-01 1.83615357e-01
-3.62062842e-01 1.72280088e-01 -1.34749576e-01 -1.24331260e+00
8.26477110e-02 -7.96275020e-01 2.17038557e-01 -9.40113127e-01
-1.12234831e+00 5.79436064e-01 -3.44953477e-01 -1.59489286e+00
-1.28273606e-01 -1.68459654e-01 -8.24143171e-01 9.28730369e-01
-1.85751927e+00 -9.46896553e-01 -9.45387840e-01 8.20107043e-01
1.04656734e-01 2.21908335e-02 8.80700588e-01 7.80668497e-01
-7.79025257e-01 7.88482606e-01 -3.66280526e-01 1.48006961e-01
1.07563782e+00 -7.64057636e-01 -3.14168721e-01 5.84832668e-01
-4.09848988e-01 3.37457776e-01 4.77640718e-01 -2.09787935e-01
-1.20802784e+00 -7.72437572e-01 6.93281651e-01 5.03550805e-02
2.30256498e-01 -8.00221339e-02 -6.77288592e-01 4.52982225e-02
2.05277324e-01 5.14118671e-01 6.32354259e-01 1.41592428e-01
6.94320723e-03 -8.46916735e-01 -1.08756959e+00 3.24075311e-01
9.83128607e-01 -4.49842632e-01 -5.25242165e-02 2.89818764e-01
2.35276401e-01 -4.30411100e-01 -1.10053015e+00 5.78645051e-01
1.03014219e+00 -1.12850153e+00 7.42306232e-01 -2.84306586e-01
3.88861060e-01 -2.43945137e-01 5.41104257e-01 -1.05892611e+00
-1.63102344e-01 -1.25883007e+00 -4.00432199e-02 1.33518517e+00
2.29039323e-02 -8.19154501e-01 1.13768983e+00 4.79571044e-01
-8.31042752e-02 -8.84200215e-01 -8.90026808e-01 -7.03039706e-01
-3.03882718e-01 -1.56313702e-01 2.24379390e-01 1.28912473e+00
1.67459548e-01 3.82868737e-01 -1.09679163e+00 -1.92905456e-01
7.19976485e-01 -1.63990501e-02 6.37962043e-01 -1.35554385e+00
-1.22713022e-01 -2.18609780e-01 -3.72134179e-01 -5.98296344e-01
-7.13551790e-02 -5.70801675e-01 -2.83165369e-04 -1.05504072e+00
1.82169259e-01 -5.29874802e-01 -5.66981971e-01 5.52663147e-01
-8.06648508e-02 7.39175439e-01 -5.11997938e-03 2.04129383e-01
-3.24516058e-01 6.00030065e-01 1.58423543e+00 2.30958387e-01
-4.24316764e-01 -2.70247441e-02 -6.38929725e-01 6.26631737e-01
8.95100951e-01 -4.70626473e-01 -6.60554767e-01 1.48740515e-01
2.49891862e-01 8.65656316e-01 2.86736310e-01 -1.14456010e+00
1.90121233e-01 1.51617259e-01 4.04169470e-01 -3.51497531e-01
3.34357440e-01 -6.85749531e-01 3.30835223e-01 3.74419928e-01
-9.73845869e-02 -6.43292665e-02 -6.74003512e-02 2.39213705e-01
-3.16679031e-01 -1.33770868e-01 9.60343659e-01 -2.31919959e-01
-3.78801644e-01 7.41575897e-01 1.58949032e-01 1.80643335e-01
9.51424360e-01 -4.83448088e-01 -4.10708547e-01 -5.34732223e-01
-8.22999656e-01 -1.07220985e-01 -1.88422367e-01 5.33827618e-02
6.20097935e-01 -1.35302007e+00 -8.68876457e-01 2.11498648e-01
-5.45952469e-02 -4.15290117e-01 9.20463920e-01 1.56896710e+00
-5.04717231e-01 1.16939999e-01 -4.80533123e-01 -5.71898639e-01
-1.23168528e+00 2.89817870e-01 7.38114417e-01 -8.35330263e-02
-7.41354942e-01 5.59234202e-01 -1.41909391e-01 -1.23574778e-01
1.99003592e-01 3.64236683e-02 -5.00772417e-01 3.09457421e-01
5.50103545e-01 4.57266152e-01 -1.00329332e-02 -2.16236860e-01
-2.38950029e-01 1.07677674e+00 1.28765613e-01 3.15804303e-01
1.11205959e+00 -3.03047359e-01 -1.07001610e-01 3.51934493e-01
1.39696348e+00 -2.55498797e-01 -1.13433897e+00 8.84586871e-02
-8.26681376e-01 -3.35424930e-01 2.74186701e-01 -7.72646785e-01
-1.62378728e+00 1.14385593e+00 1.22303915e+00 1.31207183e-02
1.60349047e+00 -6.48299754e-01 9.78600979e-01 1.71218500e-01
1.35156751e-01 -1.04204357e+00 8.00421089e-02 -1.02045283e-01
6.80655599e-01 -1.19668043e+00 1.66686445e-01 -3.01616609e-01
-6.01451039e-01 1.41511428e+00 3.83601457e-01 1.59103096e-01
7.39846647e-01 1.99704040e-02 5.59670329e-01 -2.28319783e-02
-5.15667558e-01 -1.69683434e-02 -1.06744401e-01 6.78847075e-01
4.45285857e-01 -2.52438903e-01 -6.91584826e-01 3.90394032e-01
2.06699044e-01 8.18808675e-01 5.40837526e-01 3.89679462e-01
-1.11641780e-01 -7.60775030e-01 2.99934857e-02 3.23658496e-01
-6.44447267e-01 -1.39112532e-01 3.45705122e-01 8.32219481e-01
1.74976334e-01 7.57020295e-01 -3.80867660e-01 -4.83610690e-01
2.07555890e-01 1.06937051e-01 3.18642110e-01 1.15053803e-02
-5.87865233e-01 1.44346818e-01 1.15673423e-01 -5.39420187e-01
-7.80862033e-01 -4.70372915e-01 -9.41621423e-01 -4.51747775e-01
-3.85255456e-01 -8.96886513e-02 6.47939861e-01 7.21241474e-01
4.34459925e-01 3.47638071e-01 1.07611418e+00 -7.58326769e-01
-4.89715487e-01 -1.06261599e+00 -7.46324718e-01 3.12434316e-01
3.01771313e-01 -3.31413746e-01 -5.06636024e-01 1.15961485e-01] | [13.896801948547363, 2.730717897415161] |
8a079845-31f7-4a5e-894e-15557e77c214 | performance-of-data-driven-inner-speech | 2306.10854 | null | https://arxiv.org/abs/2306.10854v1 | https://arxiv.org/pdf/2306.10854v1.pdf | Performance of data-driven inner speech decoding with same-task EEG-fMRI data fusion and bimodal models | Decoding inner speech from the brain signal via hybridisation of fMRI and EEG data is explored to investigate the performance benefits over unimodal models. Two different bimodal fusion approaches are examined: concatenation of probability vectors output from unimodal fMRI and EEG machine learning models, and data fusion with feature engineering. Same task inner speech data are recorded from four participants, and different processing strategies are compared and contrasted to previously-employed hybridisation methods. Data across participants are discovered to encode different underlying structures, which results in varying decoding performances between subject-dependent fusion models. Decoding performance is demonstrated as improved when pursuing bimodal fMRI-EEG fusion strategies, if the data show underlying structure. | ['Benjamin Metcalfe', "Eamonn O'Neill", 'Marcus Liwicki', 'Michael J. Proulx', 'Mohammad Golbabaee', 'Xi Chen', 'Oliver Watts', 'Johan Eriksson', 'Sumit Rakesh', 'Nosheen Abid', 'Kanjar De', 'Rajkumar Saini', 'Vibha Gupta', 'Foteini Simistira Liwicki', 'Scott Wellington', 'Holly Wilson'] | 2023-06-19 | null | null | null | null | ['feature-engineering'] | ['methodology'] | [ 5.96361339e-01 1.74702495e-01 5.64410806e-01 -5.59147358e-01
-1.08759773e+00 -1.97829157e-01 9.27925110e-01 8.41942430e-02
-5.28581977e-01 8.91768456e-01 6.39498174e-01 -8.62593651e-02
-4.46243256e-01 2.07886904e-01 -3.23907286e-01 -8.60083818e-01
-2.46270612e-01 7.39729777e-02 -4.15263772e-01 4.03177850e-02
4.20535803e-01 1.28265783e-01 -1.64131677e+00 8.77435029e-01
8.60776365e-01 1.17350578e+00 2.20252171e-01 8.07188809e-01
1.26622140e-01 2.27496520e-01 -7.72855163e-01 -3.39583188e-01
-1.59911849e-02 -4.05430943e-01 -5.19896984e-01 -1.31611258e-01
2.17495441e-01 2.00537488e-01 -5.23522139e-01 1.10888684e+00
8.05040061e-01 8.35615024e-02 8.06051433e-01 -1.11752248e+00
-3.81857663e-01 8.86044979e-01 -3.62049937e-01 6.20041490e-01
8.36142361e-01 8.31916928e-02 4.49259669e-01 -8.21734667e-01
-1.48266007e-03 1.33877063e+00 5.33319771e-01 3.17781478e-01
-1.85527897e+00 -8.81780207e-01 -2.60243744e-01 2.88092524e-01
-1.67825270e+00 -1.06889844e+00 4.86942768e-01 -5.38260162e-01
1.42647004e+00 3.29211295e-01 6.99369252e-01 1.37759042e+00
1.02104115e+00 5.47904015e-01 1.64390373e+00 -1.74626812e-01
3.50844651e-01 1.93726450e-01 3.50589067e-01 3.60397412e-03
1.92944542e-01 5.09626329e-01 -1.19634390e+00 -4.14585561e-01
2.69412369e-01 -4.24167812e-01 -4.64259237e-01 3.42657566e-01
-1.61125457e+00 2.82993197e-01 -3.91265713e-02 6.28504336e-01
-9.05869186e-01 -8.04605931e-02 3.31098288e-01 4.30363506e-01
3.85559440e-01 3.61197501e-01 -3.83096367e-01 -4.06120330e-01
-1.39081037e+00 9.68523473e-02 7.38562286e-01 6.92538202e-01
5.61911762e-01 2.97021508e-01 -3.86833549e-01 8.07859838e-01
5.64788043e-01 5.53930819e-01 8.53448093e-01 -8.63966048e-01
3.24281216e-01 -1.35506922e-02 -3.42821598e-01 -8.65987480e-01
-6.13122761e-01 4.02188674e-02 -5.37511766e-01 -8.51707682e-02
2.14842260e-02 -2.51272440e-01 -9.25239742e-01 1.70031083e+00
-4.93282318e-01 -3.14228795e-02 3.37137461e-01 5.66012084e-01
9.59136724e-01 5.26952505e-01 3.61457825e-01 -3.62724870e-01
1.67569959e+00 1.24570820e-02 -1.14324427e+00 -4.43548888e-01
1.15657829e-01 -5.42857111e-01 1.00251302e-01 8.67312908e-01
-1.58837187e+00 -7.99621642e-01 -1.07254601e+00 3.90645504e-01
-3.15365523e-01 -8.85754451e-02 5.74926317e-01 1.02656186e+00
-1.27444458e+00 4.40018654e-01 -7.99968362e-01 -1.63091421e-01
5.34898937e-01 8.99563313e-01 -7.06848502e-01 1.41935378e-01
-1.33867550e+00 1.26289666e+00 8.75318468e-01 1.64417580e-01
-7.98510969e-01 -6.34762585e-01 -9.41555977e-01 1.12266839e-01
-4.40114737e-01 -5.35381019e-01 9.54477727e-01 -8.59829545e-01
-1.35741293e+00 2.90160656e-01 -4.09763455e-01 -3.48247647e-01
-1.25052795e-01 1.04908638e-01 -8.04655015e-01 2.54759461e-01
-1.66152745e-01 1.15239787e+00 1.19184685e+00 -1.11646092e+00
-6.36907458e-01 -5.55705070e-01 -7.77900696e-01 3.50044668e-01
-1.46033689e-02 9.86566097e-02 2.58260727e-01 -6.51134372e-01
2.89623648e-01 -3.65889639e-01 2.85488546e-01 -1.05831528e+00
-2.12289453e-01 -1.91285059e-01 1.21034607e-01 -1.08508742e+00
1.19388723e+00 -2.27676487e+00 3.13864499e-01 4.93873537e-01
5.15993536e-01 -3.10079813e-01 -2.20644698e-01 2.13004574e-01
-5.34886360e-01 -7.00252131e-02 -5.08530259e-01 -2.13449180e-01
2.16046154e-01 4.20333594e-02 2.25522239e-02 7.31224358e-01
3.60365331e-01 9.22995865e-01 -4.86598343e-01 -2.55684942e-01
1.66529715e-01 9.03441429e-01 -2.51478165e-01 1.89660415e-01
6.69261456e-01 5.16900718e-01 1.80957824e-01 8.33074689e-01
7.14835167e-01 4.09094781e-01 1.05195649e-01 -5.82805276e-01
7.51532242e-02 3.78940225e-01 -7.78697133e-01 1.85665691e+00
2.52944119e-02 9.09994304e-01 1.96543530e-01 -1.16770637e+00
8.50008786e-01 7.85208046e-01 2.24161610e-01 -1.07175434e+00
7.21248746e-01 1.65462092e-01 7.61490881e-01 -5.67712188e-01
3.21645617e-01 -4.73739654e-01 -1.14203908e-01 2.93182462e-01
9.15402055e-01 -2.39015386e-01 1.68103173e-01 4.73940596e-02
1.10662639e+00 -2.34882504e-01 2.36050755e-01 -6.85908556e-01
3.72587949e-01 -3.90660495e-01 1.68912664e-01 8.29516411e-01
-3.25906515e-01 3.49301010e-01 3.24789554e-01 3.20014745e-01
-6.20585382e-01 -1.34169376e+00 -6.85797215e-01 9.50061500e-01
-1.79722562e-01 -5.50226867e-01 -8.68847907e-01 -1.01264298e-01
-7.45710284e-02 9.60509300e-01 -7.93857157e-01 -5.09293795e-01
-4.22768258e-02 -1.03216827e+00 7.73694456e-01 6.06634438e-01
1.77977309e-02 -8.47900093e-01 -8.37478697e-01 2.49246344e-01
-2.16552988e-01 -1.00747573e+00 5.65798581e-02 8.63764703e-01
-7.50059426e-01 -5.59468269e-01 -7.84488678e-01 -2.73281544e-01
2.79429227e-01 -6.16031550e-02 7.75559843e-01 -6.35816872e-01
-3.34230721e-01 9.28435504e-01 -2.00975031e-01 -3.50492477e-01
-3.98928195e-01 -5.12331128e-01 4.70071346e-01 7.87505060e-02
7.12068081e-01 -6.77477837e-01 -3.97765815e-01 -9.44813341e-02
-8.19211304e-01 2.01385915e-02 9.74861979e-01 9.44867313e-01
4.75014262e-02 -8.09861999e-03 9.33594108e-01 -1.01451213e-02
1.22023046e+00 -7.21770287e-01 1.80793285e-01 3.39871377e-01
-3.58641386e-01 1.07761152e-01 -1.13066800e-01 -5.80799341e-01
-1.37872720e+00 -1.31879553e-01 3.33421938e-02 -2.35750109e-01
-7.99256384e-01 5.71490765e-01 -9.08513591e-02 -1.15437336e-01
7.41028607e-01 4.54923153e-01 1.50121003e-01 -9.71132517e-02
2.23032549e-01 1.04957068e+00 8.45111847e-01 -6.20817840e-01
-3.49857956e-01 3.92956920e-02 -6.22955501e-01 -1.01556194e+00
3.85498732e-01 -1.30773978e-02 -7.40380943e-01 -3.33157063e-01
9.81393516e-01 -9.03120935e-01 -6.94670796e-01 2.77103126e-01
-1.21834445e+00 4.82175238e-02 -3.54236402e-02 1.22075212e+00
-7.87744224e-01 2.67440379e-01 -4.08095568e-01 -1.11628628e+00
-2.01876000e-01 -1.46688974e+00 1.08572876e+00 1.38649091e-01
-8.27327371e-01 -8.81090999e-01 -5.02340943e-02 2.36295730e-01
4.13598567e-01 -2.44468808e-01 6.37880564e-01 -1.14698863e+00
3.72707844e-01 -2.01920256e-01 -1.55968130e-01 2.24467993e-01
1.43021822e-01 -3.68934005e-01 -1.55987823e+00 -1.85861170e-01
4.85360682e-01 -2.42727280e-01 8.45841587e-01 6.11845672e-01
7.55506277e-01 1.30189851e-01 -3.66585195e-01 2.37194091e-01
6.86164260e-01 7.64950037e-01 6.18393183e-01 -5.07120430e-01
1.32067069e-01 9.73720312e-01 -3.32771897e-01 3.59391272e-01
1.38443962e-01 2.46639568e-02 -7.35346377e-02 3.88053715e-01
1.02524810e-01 4.58091199e-01 8.53572130e-01 1.13387549e+00
1.41932979e-01 -1.88231617e-02 -1.01581454e+00 3.58691424e-01
-1.49159098e+00 -9.14341629e-01 -3.77480239e-02 1.97961771e+00
6.29282832e-01 1.31689325e-01 -1.54892132e-01 2.41011590e-01
6.96257114e-01 -3.24249059e-01 -4.95872200e-01 -3.98286641e-01
-4.71943170e-01 5.67964137e-01 1.09464578e-01 2.62435347e-01
-8.36478353e-01 1.73858687e-01 8.08765030e+00 9.86678600e-01
-9.81804907e-01 2.68688172e-01 4.63973582e-01 -3.93176973e-01
-2.21025929e-01 -2.39560202e-01 -3.54416132e-01 6.22035027e-01
1.73952711e+00 -3.70993435e-01 7.64984965e-01 -2.92428166e-01
-4.77855317e-02 -7.26441324e-01 -1.41620934e+00 1.67576623e+00
6.20622694e-01 -9.57651436e-01 -3.31984669e-01 9.22288746e-02
1.76726863e-01 1.10046893e-01 7.79220089e-02 2.27491394e-01
-4.26931828e-02 -1.23562276e+00 1.01126814e+00 1.03586364e+00
6.79754198e-01 -4.36424196e-01 6.92214489e-01 1.83874413e-01
-9.75648582e-01 -3.03762317e-01 4.11495045e-02 -8.17091912e-02
3.17161739e-01 3.31099421e-01 -4.14124161e-01 7.61302948e-01
7.81480730e-01 5.95893502e-01 -7.69142032e-01 8.66829634e-01
3.82388562e-01 6.08552992e-01 -4.25681204e-01 2.26944059e-01
-1.15291938e-01 5.27225435e-02 5.71977854e-01 1.59096086e+00
6.21494770e-01 2.69261628e-01 -3.56765866e-01 1.14420462e+00
4.85322207e-01 -2.52335876e-01 -8.00498962e-01 -4.28790063e-01
4.16602403e-01 1.30345428e+00 -7.82750010e-01 -4.77173835e-01
-4.10286576e-01 7.58244693e-01 -1.77800208e-01 3.98818403e-01
-7.08595455e-01 -3.09322596e-01 4.21978176e-01 -4.44835603e-01
-2.46675253e-01 -1.04025483e-01 -6.56321764e-01 -1.01290047e+00
-4.39870268e-01 -6.21846974e-01 1.65086076e-01 -1.12027693e+00
-1.59681118e+00 9.54731941e-01 6.67914629e-01 -6.10204875e-01
-3.29099655e-01 -5.64832032e-01 -5.56109548e-01 1.36067772e+00
-5.60112238e-01 -5.86264074e-01 1.15170524e-01 6.75650299e-01
4.05194432e-01 -5.00939548e-01 7.28398740e-01 3.08406591e-01
-4.73194897e-01 4.34483141e-01 -1.46181345e-01 -5.60620785e-01
6.70708299e-01 -9.16994393e-01 -8.31389576e-02 5.83078027e-01
5.46626933e-02 7.69855440e-01 6.02395713e-01 -8.18600416e-01
-1.40479851e+00 -2.56697565e-01 6.15310073e-01 -3.92984599e-01
5.85802257e-01 -5.34301996e-01 -1.13596344e+00 2.32962921e-01
1.18261456e+00 -5.61878920e-01 1.06868184e+00 -2.08956152e-01
1.21374115e-01 2.31237873e-01 -1.13381362e+00 3.35401893e-01
6.62333190e-01 -8.88206005e-01 -1.37109149e+00 -1.76741734e-01
1.55154690e-01 1.68530911e-01 -1.24556613e+00 5.47588110e-01
7.43052781e-01 -7.10234582e-01 7.93564796e-01 -5.13744116e-01
-1.21715300e-01 -2.65206750e-02 -6.87518477e-01 -1.69173646e+00
-5.17846406e-01 -5.69523811e-01 -2.46233478e-01 1.04567039e+00
4.07395512e-01 -8.55177164e-01 5.75627275e-02 8.23870301e-01
-3.47633541e-01 -3.08870375e-01 -1.41341114e+00 -4.53084826e-01
-2.80575491e-02 -8.39851797e-01 3.44808757e-01 7.10799694e-01
8.19853783e-01 5.27508140e-01 1.22203842e-01 -1.48999155e-01
4.30378586e-01 -7.25515008e-01 1.65106896e-02 -1.11208212e+00
1.16883464e-01 -9.31004524e-01 -6.76246166e-01 -4.68068808e-01
4.38524544e-01 -1.23003614e+00 3.66086721e-01 -1.26897228e+00
1.07742511e-01 2.33780578e-01 -6.34467602e-01 5.49113750e-01
-2.79553365e-02 2.15214372e-01 1.00407943e-01 -5.91743998e-02
-1.38469338e-01 6.28743112e-01 4.06301320e-01 -1.63487777e-01
-6.33651614e-02 -6.87749088e-01 -8.71617079e-01 4.60934192e-01
7.36110210e-01 -2.03257278e-01 -4.64725405e-01 -1.76899597e-01
2.13498678e-02 6.19470417e-01 4.09741968e-01 -1.42730999e+00
4.57822174e-01 4.36452806e-01 9.65151250e-01 -5.34908116e-01
6.14444315e-01 -6.58133388e-01 4.87594306e-01 1.83438331e-01
-4.05144423e-01 2.27867275e-01 7.01039135e-01 4.38306749e-01
-2.87727535e-01 1.44013399e-02 5.94468296e-01 2.20899612e-01
-5.34169316e-01 -4.36843246e-01 -1.06539750e+00 -3.17485452e-01
8.81345868e-01 -6.83892787e-01 -2.25306064e-01 -1.43370584e-01
-1.48226416e+00 -1.48812935e-01 -5.24500310e-01 2.82574445e-01
8.83938193e-01 -1.38425636e+00 -8.21161330e-01 8.19064856e-01
-2.29379162e-01 -8.09841871e-01 5.01749754e-01 1.74540901e+00
3.01896900e-01 5.80825865e-01 -6.95756972e-01 -8.33152413e-01
-9.61337686e-01 3.58326018e-01 1.96150988e-01 5.53811431e-01
-3.46472144e-01 8.18020284e-01 4.44569886e-01 1.02473320e-02
9.81516093e-02 -2.40696967e-01 -1.70983389e-01 5.85084200e-01
7.10843205e-01 5.17945349e-01 5.24387956e-01 -8.99054110e-01
-4.62227285e-01 -1.35486051e-01 -1.19199846e-02 -7.22040117e-01
1.12572086e+00 -3.15817624e-01 -1.08573236e-01 7.86183417e-01
9.46676672e-01 -5.74483573e-01 -9.76971030e-01 -4.61773452e-04
-9.00960863e-02 -1.72284082e-01 3.60372573e-01 -1.09562528e+00
-7.64203310e-01 1.12222290e+00 1.03268063e+00 4.01962608e-01
1.14269304e+00 1.10520482e-01 1.80132657e-01 2.36133397e-01
2.20614493e-01 -1.15827823e+00 -4.65336204e-01 2.78198123e-01
1.04097784e+00 -9.10421431e-01 -4.41173315e-02 2.76960790e-01
-9.04520571e-01 1.20389390e+00 1.82075188e-01 -1.06054537e-01
1.00418687e+00 6.47645414e-01 -5.43170094e-01 -3.96690428e-01
-9.10211682e-01 -1.33605614e-01 6.63349092e-01 6.86131001e-01
7.01769948e-01 2.47773260e-01 -4.70673382e-01 1.22753000e+00
-2.86622405e-01 -1.34493530e-01 3.36026363e-02 1.03676295e+00
-2.46467650e-01 -5.54752409e-01 -9.15070474e-01 1.13827074e+00
-3.19551200e-01 -3.37215453e-01 -3.38738889e-01 5.75922847e-01
2.49048114e-01 1.18671799e+00 3.46374393e-01 -7.57475138e-01
4.47998494e-02 7.98334777e-01 8.16000521e-01 -2.26177275e-01
-8.19333136e-01 5.53532124e-01 6.82716593e-02 -5.29144108e-01
-6.17163658e-01 -1.12292922e+00 -1.16632009e+00 -3.20300385e-02
-4.64193851e-01 2.16120198e-01 7.61834502e-01 1.25117505e+00
4.96954709e-01 9.56214666e-01 -1.06216617e-01 -1.45941043e+00
-7.11550191e-02 -1.42720914e+00 -6.21243000e-01 -5.16681857e-02
2.05288440e-01 -1.05180287e+00 -2.91721106e-01 3.26407224e-01] | [13.007132530212402, 3.43220853805542] |
b1d2034a-8eda-4306-b357-041b4d2736e1 | quantifying-the-robustness-of-deep | 2305.11347 | null | https://arxiv.org/abs/2305.11347v1 | https://arxiv.org/pdf/2305.11347v1.pdf | Quantifying the robustness of deep multispectral segmentation models against natural perturbations and data poisoning | In overhead image segmentation tasks, including additional spectral bands beyond the traditional RGB channels can improve model performance. However, it is still unclear how incorporating this additional data impacts model robustness to adversarial attacks and natural perturbations. For adversarial robustness, the additional information could improve the model's ability to distinguish malicious inputs, or simply provide new attack avenues and vulnerabilities. For natural perturbations, the additional information could better inform model decisions and weaken perturbation effects or have no significant influence at all. In this work, we seek to characterize the performance and robustness of a multispectral (RGB and near infrared) image segmentation model subjected to adversarial attacks and natural perturbations. While existing adversarial and natural robustness research has focused primarily on digital perturbations, we prioritize on creating realistic perturbations designed with physical world conditions in mind. For adversarial robustness, we focus on data poisoning attacks whereas for natural robustness, we focus on extending ImageNet-C common corruptions for fog and snow that coherently and self-consistently perturbs the input data. Overall, we find both RGB and multispectral models are vulnerable to data poisoning attacks regardless of input or fusion architectures and that while physically realizable natural perturbations still degrade model performance, the impact differs based on fusion architecture and input data. | ['Eleanor Byler', 'Myles Mckay', 'Charles Godfrey', 'Elise Bishoff'] | 2023-05-18 | null | null | null | null | ['data-poisoning'] | ['adversarial'] | [ 5.20433724e-01 -1.77036032e-01 3.78194213e-01 1.39151484e-01
-4.73917037e-01 -1.30953026e+00 5.33379376e-01 1.45466119e-01
-3.62553328e-01 5.14492273e-01 -8.38764906e-02 -7.25576520e-01
-2.29591993e-03 -9.46514368e-01 -8.57112348e-01 -9.71777916e-01
-3.16303619e-03 -8.54405388e-02 2.85305351e-01 -5.85821450e-01
1.37526318e-02 9.24653709e-01 -1.33426428e+00 -1.25934333e-02
9.58105922e-01 7.43689239e-01 -6.33082509e-01 1.22653699e+00
5.48360646e-01 8.76559556e-01 -1.06018937e+00 -2.44578883e-01
9.33926463e-01 -1.92131490e-01 -8.20599258e-01 4.87727905e-03
5.97711384e-01 -5.48759162e-01 -6.47670209e-01 1.25009644e+00
6.72662735e-01 1.83567792e-01 2.83337206e-01 -1.52589536e+00
-4.55300421e-01 4.63729084e-01 -3.08247328e-01 3.05631489e-01
4.49148081e-02 1.26631105e+00 3.89262766e-01 2.69248843e-01
3.18379819e-01 1.11501658e+00 9.38358009e-01 5.86881876e-01
-1.40893650e+00 -7.85668790e-01 2.11743563e-01 -1.88823685e-01
-1.16908550e+00 -4.04859573e-01 5.71248114e-01 -2.56791115e-01
9.22405839e-01 7.63592243e-01 5.45188844e-01 1.12942100e+00
4.56958115e-01 1.20329544e-01 1.40404963e+00 7.04394504e-02
3.19877893e-01 -2.08959565e-01 1.24383122e-02 2.70467252e-01
4.22187835e-01 6.40377760e-01 -2.51088411e-01 -5.67117333e-01
3.55665505e-01 -4.09580559e-01 -5.10754764e-01 2.59188384e-01
-8.04270148e-01 5.43996453e-01 8.76938879e-01 -1.15539692e-01
-3.57671678e-01 5.51627159e-01 2.49925017e-01 3.07399869e-01
1.68050662e-01 8.20425630e-01 -4.90992904e-01 1.30194426e-01
-7.61707008e-01 5.15446603e-01 5.71053505e-01 5.21391869e-01
8.69137347e-01 5.88179171e-01 2.67281700e-02 1.18018068e-01
1.79681495e-01 1.14271200e+00 -1.39845744e-01 -1.11762869e+00
2.56624013e-01 2.22707599e-01 1.09403253e-01 -1.15477097e+00
-6.25431120e-01 -2.39291027e-01 -4.83355045e-01 7.27403760e-01
4.48841333e-01 -5.57567358e-01 -1.57915938e+00 1.65588677e+00
2.93455452e-01 3.74442369e-01 4.18496609e-01 1.06106496e+00
7.27003038e-01 4.39947844e-01 3.26859653e-01 2.67631084e-01
9.48748410e-01 -1.88820422e-01 -4.68840927e-01 -5.94884515e-01
2.99265504e-01 -8.64835203e-01 9.19782937e-01 2.38901585e-01
-7.94164836e-01 -1.72880143e-01 -1.20387995e+00 3.60591829e-01
-8.61016929e-01 -1.05217397e+00 4.81378138e-01 1.25268984e+00
-8.94271255e-01 5.95625460e-01 -9.25850332e-01 -3.59383792e-01
4.08542514e-01 5.14023364e-01 -2.30081216e-01 -1.80759579e-02
-1.54310966e+00 1.20280242e+00 2.50704855e-01 1.43569440e-01
-9.49543357e-01 -8.87915909e-01 -8.00570428e-01 -3.77222657e-01
1.16715536e-01 -6.25768423e-01 8.10799837e-01 -9.83609378e-01
-1.20863450e+00 4.53383774e-01 3.53551477e-01 -8.71848226e-01
5.84522963e-01 1.59389094e-01 -6.58823490e-01 3.52927536e-01
-2.69793600e-01 8.68626595e-01 9.67927456e-01 -1.80787861e+00
-1.66208982e-01 -2.47664317e-01 6.21959329e-01 2.48167008e-01
-1.37002066e-01 7.03220302e-03 -2.28329357e-02 -7.31583714e-01
-4.90584113e-02 -1.25013149e+00 -4.43387628e-01 -6.13723956e-02
-6.77735925e-01 1.20030653e+00 1.21115768e+00 -6.04085445e-01
6.69421673e-01 -2.12462282e+00 -1.96674243e-01 5.55494249e-01
-1.17883831e-02 6.21553302e-01 -1.76927522e-01 1.24397658e-01
3.57801393e-02 6.84235990e-01 -7.34068871e-01 9.23922658e-02
-1.87245712e-01 5.76088309e-01 -6.02876544e-01 8.23796690e-01
2.56884664e-01 8.07968080e-01 -6.74263358e-01 -1.79718025e-02
4.71226901e-01 4.96764004e-01 -4.70045298e-01 -1.65866330e-01
-1.13634057e-01 6.61330462e-01 -1.49406061e-01 9.13153589e-01
1.13460743e+00 3.63167584e-01 -4.46534663e-01 -2.57368118e-01
2.41377041e-01 -1.53814763e-01 -1.09713256e+00 8.73282433e-01
-1.97387308e-01 3.95624906e-01 5.36445320e-01 -1.59412906e-01
4.18498188e-01 1.25925347e-01 2.99078375e-01 -5.47365189e-01
1.87092572e-01 -3.18214148e-02 3.15366626e-01 -2.24073350e-01
7.16874063e-01 -2.94092238e-01 -1.71981245e-01 4.60549265e-01
-4.91608262e-01 -7.38148212e-01 -2.99133688e-01 2.51904130e-01
1.42036629e+00 -1.59069225e-01 -2.84899026e-01 -8.27820674e-02
1.03087954e-01 5.56815624e-01 5.35972238e-01 1.13971496e+00
-5.55615067e-01 7.15275526e-01 1.50905982e-01 -2.96914671e-02
-6.65136874e-01 -1.33572710e+00 -1.86320454e-01 6.61410153e-01
6.48844779e-01 -5.33364601e-02 -8.32699120e-01 -7.05983222e-01
5.14127851e-01 8.76465082e-01 -5.71725607e-01 -7.72485435e-01
-2.19175950e-01 -1.18879175e+00 1.64027035e+00 3.32647562e-01
8.40943694e-01 -7.96563804e-01 -6.16533041e-01 1.11336512e-02
-1.29657134e-01 -1.25744879e+00 -8.01173598e-02 3.81534815e-01
-8.08124304e-01 -1.28574371e+00 5.08870510e-03 3.05585116e-01
5.21414757e-01 3.32055151e-01 9.10375476e-01 3.02561045e-01
-3.00352186e-01 8.07185352e-01 -5.23433030e-01 -8.16584051e-01
-9.14326310e-01 -1.77201852e-01 1.77071601e-01 -2.02067941e-01
-5.67289516e-02 -5.43105423e-01 -3.93361330e-01 3.12091202e-01
-1.51548886e+00 -6.00822270e-01 -2.78964508e-02 3.41246903e-01
1.69099584e-01 2.93197036e-01 6.33424241e-03 -7.71660507e-01
4.58456159e-01 -4.85137969e-01 -4.14729714e-01 -6.16493300e-02
-4.54304159e-01 -1.68254107e-01 7.94362664e-01 -3.90120685e-01
-8.02893043e-01 2.98786443e-03 -1.57136451e-02 -5.79274774e-01
-5.39009571e-01 3.70844901e-01 -3.04422140e-01 -9.52221155e-01
1.17056143e+00 -1.88800246e-02 -1.39279574e-01 3.53002585e-02
4.24350351e-01 4.11309838e-01 7.07621098e-01 -6.14968777e-01
1.37610030e+00 8.88830364e-01 1.07010596e-01 -8.53821337e-01
-2.61581242e-01 3.05311363e-02 -2.55788445e-01 -3.75412315e-01
8.20405185e-01 -8.41973841e-01 -4.79989260e-01 1.14820516e+00
-7.54408300e-01 -7.49315858e-01 -2.80606806e-01 8.01536161e-03
-2.36283526e-01 4.02139425e-01 -5.30488253e-01 -5.75978279e-01
-1.22041047e-01 -1.39691937e+00 6.98115528e-01 1.76850393e-01
-4.35636528e-02 -9.16307151e-01 -2.72438586e-01 4.76222754e-01
8.02228153e-01 9.64878142e-01 5.67131519e-01 -4.27923888e-01
-6.70071304e-01 -2.09775776e-01 2.12830976e-02 3.38305980e-01
5.02934158e-01 3.77135456e-01 -1.37876153e+00 -4.48672265e-01
4.01478037e-02 -3.40084821e-01 9.07682776e-01 1.48769215e-01
8.14710438e-01 -4.41319168e-01 1.82989091e-01 1.08522987e+00
1.40944099e+00 -1.33352041e-01 9.88882542e-01 6.68301284e-01
9.32870150e-01 4.12759542e-01 4.15747553e-01 2.26832211e-01
7.50416368e-02 1.38283685e-01 1.23431587e+00 -5.56140602e-01
6.49120659e-02 3.31948072e-01 3.85941088e-01 -2.99291551e-01
-2.26338836e-03 -4.00459051e-01 -1.10383523e+00 2.34218135e-01
-1.43641317e+00 -1.03228128e+00 -2.13721111e-01 2.04371428e+00
7.36654282e-01 3.28303814e-01 -5.98860122e-02 2.06471711e-01
6.98860884e-01 6.07661903e-01 -1.03610849e+00 -5.70869088e-01
-6.44069135e-01 2.13422328e-01 1.55929995e+00 6.37765169e-01
-1.25847733e+00 1.09525967e+00 6.90102577e+00 2.09169626e-01
-1.43797159e+00 -1.38478845e-01 5.82233369e-01 -4.31106001e-01
-3.56748074e-01 2.15044320e-02 -3.07846755e-01 2.99273759e-01
1.00212777e+00 5.12631536e-02 7.74271309e-01 2.71376848e-01
4.25340593e-01 -3.52182806e-01 -4.42807615e-01 3.69327962e-01
-7.26446062e-02 -1.21047652e+00 1.06325597e-01 2.65822262e-01
8.03379118e-01 4.03072506e-01 3.18288863e-01 -5.10576367e-02
9.86124873e-01 -1.11747718e+00 6.30115390e-01 3.18624020e-01
4.67416227e-01 -8.38504434e-01 6.02794409e-01 2.87325289e-02
-8.00140440e-01 -1.04967177e-01 1.39838792e-02 -1.66438207e-01
-4.81389202e-02 1.94926605e-01 -4.57041502e-01 5.96838415e-01
8.87960196e-01 1.95186451e-01 -8.93185973e-01 7.36745417e-01
-2.70536214e-01 9.99807239e-01 -7.06668675e-01 7.34383941e-01
4.71970737e-01 2.50771433e-01 9.02189076e-01 1.04320586e+00
-2.08218098e-01 3.30217659e-01 1.36648342e-01 6.67493701e-01
1.14917800e-01 -6.55120611e-01 -8.97639394e-01 1.04737535e-01
5.40842593e-01 8.61501634e-01 -8.12329412e-01 -5.13112508e-02
-3.01071908e-02 9.73352730e-01 -4.00276184e-01 7.36575663e-01
-1.14114571e+00 -2.63625026e-01 1.59174907e+00 -9.00119636e-03
1.00126296e-01 -3.18239629e-01 -7.25457609e-01 -9.33506787e-01
-2.24919498e-01 -1.47096741e+00 3.13917875e-01 -8.01631093e-01
-1.13794029e+00 2.81546474e-01 -1.41943777e-02 -1.06126308e+00
1.52971640e-01 -5.71208954e-01 -8.51092398e-01 1.03850138e+00
-1.44302678e+00 -1.42562330e+00 -3.97690386e-01 1.02138913e+00
-3.28088373e-01 1.91945940e-01 7.46176183e-01 -2.41356686e-01
-4.86591876e-01 6.81039095e-01 5.44099472e-02 1.10396422e-01
8.14300299e-01 -1.20777452e+00 7.16119945e-01 1.66644406e+00
-3.97260308e-01 6.90546393e-01 1.25802267e+00 -9.28316832e-01
-1.49442625e+00 -1.49638128e+00 -3.45755666e-01 -5.95003188e-01
7.20206439e-01 9.03314576e-02 -9.71985817e-01 6.26303792e-01
1.09998651e-01 3.34050953e-01 4.94501948e-01 -4.71082896e-01
-5.99505901e-01 -1.20757848e-01 -1.85212600e+00 7.90824294e-01
6.08507752e-01 -7.61479378e-01 -1.10634357e-01 3.65802407e-01
9.74373639e-01 -8.16572309e-01 -9.58003283e-01 5.94824970e-01
6.62425160e-02 -1.01824021e+00 1.24452806e+00 -6.52424693e-01
-1.98425613e-02 -8.25417995e-01 -4.29401129e-01 -1.42104721e+00
-1.13106236e-01 -7.11056590e-01 5.98922558e-02 1.18270886e+00
2.93541580e-01 -1.07115567e+00 5.84158957e-01 1.07663977e+00
-1.41166389e-01 -1.01029880e-01 -7.78833926e-01 -8.48287046e-01
4.37639475e-01 -7.88564682e-01 9.83793318e-01 1.06154239e+00
-5.98141849e-01 -7.14709044e-01 -1.46379054e-01 1.36266983e+00
6.84520483e-01 -3.66850555e-01 9.37754750e-01 -4.94445860e-01
-2.68552989e-01 -2.53091574e-01 -7.00881362e-01 -2.43112464e-02
-7.04454258e-02 -4.25766110e-01 1.37219980e-01 -1.16896164e+00
-5.62671542e-01 -4.47472990e-01 -3.00777763e-01 8.74230325e-01
-5.60103118e-01 7.18560398e-01 5.36759436e-01 2.06035659e-01
2.94087864e-02 1.38919905e-01 8.07994485e-01 -4.98320818e-01
-1.64492816e-01 -1.14400446e-01 -9.10094500e-01 6.43162310e-01
1.02204359e+00 -5.64461350e-01 -2.13826209e-01 -5.79166234e-01
-6.82819337e-02 -5.28987825e-01 9.43321526e-01 -1.29138160e+00
8.77779722e-02 -5.12217283e-01 3.22300076e-01 4.28900346e-02
2.74062634e-01 -9.26983535e-01 2.34963626e-01 6.83902442e-01
1.49957076e-01 -4.13295068e-02 1.06722474e+00 5.07941008e-01
1.14217430e-01 4.16298434e-02 1.00822008e+00 -1.95165709e-01
-6.98503673e-01 2.60927469e-01 -7.50098944e-01 1.14613831e-01
1.14692795e+00 -4.28005517e-01 -9.02369142e-01 -3.74849707e-01
-7.10127413e-01 2.71182418e-01 1.28777814e+00 2.36469984e-01
1.99023485e-01 -7.81132400e-01 -5.87918103e-01 1.30549252e-01
-6.63647079e-05 -7.04157799e-02 2.13740289e-01 3.71162742e-01
-8.35036218e-01 -3.62798095e-01 -2.45678321e-01 -5.20906329e-01
-1.32702029e+00 4.31274712e-01 9.80049610e-01 1.71553865e-01
-1.04444452e-01 8.71209443e-01 -3.62289488e-01 -7.13147163e-01
8.50005299e-02 -2.96951890e-01 3.76926094e-01 -2.01731324e-01
3.12141806e-01 4.04630810e-01 3.42420429e-01 -9.22710299e-01
-5.00719190e-01 3.12460154e-01 1.63921043e-01 -1.89763084e-01
1.00773704e+00 -5.80072291e-02 -3.36901881e-02 -5.84259629e-03
6.18289173e-01 3.60148028e-02 -1.53712213e+00 6.21209703e-02
-6.95007861e-01 -6.62511051e-01 3.03966105e-01 -1.09111071e+00
-1.28251898e+00 5.90717852e-01 8.14509511e-01 4.14121896e-01
1.41973066e+00 -4.44787562e-01 7.73844719e-01 4.41779673e-01
2.34056219e-01 -8.21163476e-01 -3.74487996e-01 6.04877234e-01
6.36943936e-01 -1.14844584e+00 2.81830430e-01 -3.84507179e-01
-5.81066012e-01 7.55212963e-01 7.74174452e-01 -1.02205947e-02
6.03320122e-01 7.40635455e-01 6.07936442e-01 -5.31765111e-02
-6.65180683e-02 -3.11859846e-01 -2.20046237e-01 1.10196030e+00
-3.24240416e-01 2.08499772e-03 5.69600284e-01 -5.12629300e-02
-3.63571763e-01 -3.83051693e-01 8.62157822e-01 1.25728905e+00
-2.73760140e-01 -7.65649617e-01 -1.26968110e+00 4.76409912e-01
-4.08188820e-01 -2.64344901e-01 -7.74717212e-01 8.62383962e-01
3.35477054e-01 1.42670012e+00 -1.22233950e-01 -6.77983999e-01
3.91543031e-01 -1.53909728e-01 3.07075292e-01 -2.80750811e-01
-1.32504344e+00 -4.84655678e-01 9.72559601e-02 -7.22210228e-01
-3.34458947e-01 -7.18002498e-01 -1.34286392e+00 -9.80661273e-01
-4.28659499e-01 -3.37826848e-01 6.04469597e-01 1.01713860e+00
3.63086820e-01 7.10237622e-01 5.53433061e-01 -1.16327608e+00
-4.75189179e-01 -5.38318455e-01 -3.36603284e-01 3.50695759e-01
5.76218069e-01 -2.02810541e-01 -7.62174249e-01 -5.25973104e-02] | [5.477110862731934, 7.953195095062256] |
7af493b7-9d06-487d-bc77-9e51fe416e5d | cyber-risk-frequency-severity-and-insurance | 2111.03366 | null | https://arxiv.org/abs/2111.03366v2 | https://arxiv.org/pdf/2111.03366v2.pdf | Cyber Risk Frequency, Severity and Insurance Viability | In this study an exploration of insurance risk transfer is undertaken for the cyber insurance industry in the United States of America, based on the leading industry dataset of cyber events provided by Advisen. We seek to address two core unresolved questions. First, what factors are the most significant covariates that may explain the frequency and severity of cyber loss events and are they heterogeneous over cyber risk categories? Second, is cyber risk insurable in regards to the required premiums, risk pool sizes and how would this decision vary with the insured companies industry sector and size? We address these questions through a combination of regression models based on the class of Generalised Additive Models for Location Shape and Scale (GAMLSS) and a class of ordinal regressions. These models will then form the basis for our analysis of frequency and severity of cyber risk loss processes. We investigate the viability of insurance for cyber risk using a utility modelling framework with premium calculated by classical certainty equivalence analysis utilising the developed regression models. Our results provide several new key insights into the nature of insurability of cyber risk and rigorously address the two insurance questions posed in a real data driven case study analysis. | ['Georgy Sofronov', 'Jiwook Jang', 'Stefan Trück', 'Pavel V. Shevchenko', 'Gareth W. Peters', 'Matteo Malavasi'] | 2021-11-05 | null | null | null | null | ['additive-models'] | ['methodology'] | [ 2.21207604e-01 5.57901084e-01 -2.51264840e-01 1.30342871e-01
-6.41021967e-01 -7.79334843e-01 4.45350677e-01 4.57663924e-01
-2.28657290e-01 5.70448697e-01 5.01438558e-01 -1.07360280e+00
-7.71156728e-01 -1.03807878e+00 -5.43863714e-01 -8.94221589e-02
1.16062261e-01 1.05430722e-01 -7.51804486e-02 -2.64341205e-01
3.63999188e-01 3.21713567e-01 -1.61895275e+00 -3.29071023e-02
8.81556630e-01 1.17218828e+00 -1.61305785e-01 1.35742173e-01
3.27085882e-01 7.20631480e-01 -4.34203774e-01 -7.67392993e-01
3.82040232e-01 -7.44079500e-02 -5.05986035e-01 -4.03950751e-01
-3.79771948e-01 -3.75035077e-01 4.05316681e-01 7.66528606e-01
1.67871967e-01 -4.23659205e-01 9.57050622e-01 -1.26969194e+00
-5.29333949e-01 6.77453637e-01 1.66170709e-02 -6.20891191e-02
7.02586770e-01 8.49335361e-03 9.29436684e-01 -2.33040452e-02
5.97047031e-01 1.10611343e+00 6.75048292e-01 -2.04262227e-01
-1.24577677e+00 -5.41273832e-01 1.19593866e-01 -2.09695786e-01
-9.14669633e-01 1.95912257e-01 4.26725686e-01 -8.97846937e-01
8.94409955e-01 5.09618342e-01 4.87944156e-01 8.24658573e-01
6.70349300e-01 -2.08295867e-01 1.32320714e+00 -6.44293368e-01
2.73681968e-01 5.68267405e-01 3.12607557e-01 -4.50692624e-02
6.91747129e-01 7.04216599e-01 -4.42043357e-02 -6.00067675e-01
7.01495528e-01 1.51997134e-01 1.36696801e-01 -5.09914830e-02
-1.03762221e+00 9.34733808e-01 9.65510681e-02 2.64697164e-01
-6.60836756e-01 -2.91844853e-03 2.44675949e-01 7.08254993e-01
4.44602638e-01 2.62453169e-01 -6.95167363e-01 -4.62697297e-02
-1.75394356e-01 2.86784410e-01 1.07329905e+00 6.81283772e-01
3.36231470e-01 -3.21612030e-01 -1.64966471e-02 2.76715845e-01
6.52394235e-01 2.21770421e-01 -1.18182600e-01 -9.81624961e-01
7.51352429e-01 6.93893075e-01 3.55631977e-01 -6.34818316e-01
4.62634563e-02 -8.48735049e-02 -2.08027929e-01 7.43421674e-01
5.38703442e-01 -5.63777089e-01 -4.57355082e-01 1.55214167e+00
-5.25784083e-02 -2.73811340e-01 6.51700571e-02 1.93430111e-01
-1.45508409e-01 4.92158532e-02 5.79423726e-01 -2.81916142e-01
1.71990061e+00 1.73801005e-01 -5.42361140e-01 1.79749399e-01
5.92668533e-01 -3.82116139e-01 1.00023329e+00 5.04776239e-01
-1.10098183e+00 -1.99185103e-01 -8.98816228e-01 5.13463140e-01
-4.74547476e-01 -5.43317020e-01 7.39621878e-01 1.51341271e+00
-7.01569498e-01 7.14434803e-01 -3.31879586e-01 -1.61439091e-01
2.45000944e-01 -8.00972506e-02 1.64307952e-01 1.75585821e-01
-1.53587961e+00 1.25179708e+00 -1.04123242e-01 -3.73371452e-01
-5.18792033e-01 -9.48178649e-01 -5.62551439e-01 1.70062006e-01
2.22852811e-01 -5.99213600e-01 9.70205188e-01 -7.46304572e-01
-8.44314933e-01 4.97346759e-01 6.73207998e-01 -6.84954762e-01
7.96722412e-01 -8.40013176e-02 -5.41264892e-01 -2.39330962e-01
1.46119699e-01 -3.75652909e-01 1.49504915e-01 -1.08216190e+00
-5.54170787e-01 -4.56019968e-01 4.40310329e-01 -1.05959207e-01
1.41964272e-01 5.28210640e-01 6.83974028e-01 -9.09294844e-01
-2.88249940e-01 -7.69878805e-01 -3.73726785e-01 -5.64718604e-01
6.14223117e-03 5.86644672e-02 1.08910939e-02 -9.43569005e-01
1.29538774e+00 -2.02778959e+00 -3.41321111e-01 3.29664260e-01
-4.77512389e-01 -3.53630930e-01 5.64767778e-01 8.07420135e-01
-2.82989472e-01 5.21929920e-01 -4.06603336e-01 3.09443742e-01
4.04839337e-01 -3.12524229e-01 -5.55772662e-01 2.36457914e-01
1.96277529e-01 5.53317487e-01 -2.68682450e-01 -3.34967449e-02
3.73291820e-01 4.22476977e-01 -4.41945434e-01 4.98302169e-02
7.15091825e-02 -4.97543402e-02 -4.26085681e-01 8.94487143e-01
6.77132428e-01 3.93122375e-01 2.34631240e-01 1.26334310e-01
-4.11401838e-01 3.78154248e-01 -1.03389645e+00 9.88712370e-01
-4.48915780e-01 -1.53135464e-01 -1.24876194e-01 -6.15477860e-01
7.19113946e-01 5.25442481e-01 2.92725474e-01 -4.25387710e-01
1.80390939e-01 5.47938310e-02 2.22902168e-02 -1.22896835e-01
-8.96287933e-02 -8.05799067e-01 -4.75233406e-01 6.75195932e-01
-2.13974401e-01 -9.32937041e-02 -4.63328391e-01 -1.83935046e-01
1.31427598e+00 4.62720878e-02 3.93985331e-01 -4.78714377e-01
1.15983389e-01 -1.05777800e-01 4.85226363e-01 2.60933489e-01
-2.47150138e-01 4.10900205e-01 1.17592335e+00 -1.31423445e-02
-6.10462189e-01 -1.32145452e+00 -4.71044958e-01 4.34382349e-01
-5.75482938e-03 2.37267777e-01 -8.40460777e-01 -6.44280434e-01
4.85716730e-01 1.32572925e+00 -5.85065544e-01 -2.65183985e-01
2.82225963e-02 -9.26595867e-01 3.04747045e-01 2.97790021e-01
-7.10848868e-02 -1.15864933e+00 -1.04174316e+00 2.45619923e-01
3.12208146e-01 -6.67571962e-01 5.06832413e-02 2.36107260e-01
-8.21192443e-01 -1.34759343e+00 -3.79499882e-01 -1.10344969e-01
2.12347850e-01 -3.55941772e-01 1.12925482e+00 -6.46965131e-02
-3.24978679e-01 6.43716931e-01 -2.79726177e-01 -1.11389017e+00
-5.61896801e-01 -3.95830035e-01 -1.75201789e-01 -1.09171987e-01
3.15910280e-01 -7.44951248e-01 -5.39527416e-01 2.95944601e-01
-9.96505797e-01 -7.60665178e-01 5.82678318e-01 -4.98913974e-03
1.78586736e-01 5.65552711e-01 1.15541017e+00 -9.48421180e-01
1.19838595e+00 -1.00253904e+00 -6.80404246e-01 4.30758387e-01
-1.22754681e+00 -3.07166755e-01 -3.04786563e-01 -1.59518823e-01
-1.35314298e+00 -2.64174521e-01 -9.97556224e-02 1.90333098e-01
-2.45921806e-01 6.90353215e-01 -5.91679156e-01 1.55589476e-01
4.98824209e-01 -6.84465706e-01 1.96849778e-01 -4.71257359e-01
3.16051632e-01 4.77594703e-01 -1.13564417e-01 -5.21892071e-01
6.75082386e-01 4.01108593e-01 -1.87851846e-01 -1.42046764e-01
-1.90882802e-01 -7.30906278e-02 -3.75393420e-01 -4.85821962e-01
1.07536840e+00 -8.46483946e-01 -9.88774180e-01 2.94631898e-01
-6.87794626e-01 -3.28789055e-01 -5.23360610e-01 5.20621598e-01
-6.99253678e-01 2.14208618e-01 -5.87775111e-01 -1.42911518e+00
2.06830772e-03 -9.87263203e-01 4.19690490e-01 -3.64570647e-01
-5.12739301e-01 -1.00247455e+00 5.59923090e-02 5.66976249e-01
4.12734658e-01 1.03606772e+00 1.32073832e+00 -4.25253391e-01
-7.08542585e-01 -5.64606130e-01 4.43279603e-03 5.30702055e-01
1.64404958e-01 -2.84646153e-01 -6.73616230e-01 8.84415135e-02
6.34696841e-01 -4.40649763e-02 5.29315472e-01 6.22126937e-01
4.19585258e-01 -2.61881918e-01 -1.71903297e-01 -4.88477247e-03
1.86174369e+00 4.70581353e-01 8.72051597e-01 5.76197863e-01
1.09728254e-01 1.42254913e+00 1.11471379e+00 4.48777378e-01
3.73613507e-01 5.39529443e-01 6.59542024e-01 3.57375234e-01
5.04826844e-01 -2.15114996e-01 4.56379890e-01 6.43629134e-02
-2.98208177e-01 9.88477021e-02 -1.01202643e+00 4.81966168e-01
-1.50319052e+00 -9.11708891e-01 -1.79461598e-01 2.42128706e+00
3.63266319e-01 6.40170276e-01 5.10993361e-01 4.40485924e-01
6.19243503e-01 -1.56326801e-01 -2.12237790e-01 -7.57366359e-01
2.36605942e-01 5.57014300e-03 9.98807907e-01 3.29343379e-01
-5.95919013e-01 -1.42103685e-02 7.44338989e+00 1.95354775e-01
-6.33586571e-02 1.05712831e-01 9.30080473e-01 1.64206147e-01
-8.95184219e-01 5.92969954e-01 -4.46546584e-01 6.68987751e-01
1.40719163e+00 -4.22813654e-01 2.72165596e-01 5.93419075e-01
2.08714619e-01 -2.44364724e-01 -8.54960322e-01 -1.57558918e-01
-3.84174317e-01 -9.42420363e-01 -1.56421959e-01 7.00627744e-01
3.83163840e-01 -4.18071538e-01 2.28801683e-01 2.16168642e-01
5.82028806e-01 -8.74052644e-01 1.02866840e+00 1.03918719e+00
6.03949904e-01 -1.02477050e+00 8.17976058e-01 2.63532549e-01
-6.56746686e-01 -5.79908729e-01 1.77687705e-01 -2.57885516e-01
7.94201434e-01 7.46871352e-01 -3.26542079e-01 6.98221207e-01
6.24191225e-01 -3.74981537e-02 -8.03123042e-02 4.23218042e-01
2.57165045e-01 6.23122275e-01 4.65150364e-02 5.46027899e-01
-2.56362587e-01 -4.56203103e-01 2.38699824e-01 6.45007730e-01
6.67622924e-01 8.59594718e-02 -7.55191982e-01 1.19696963e+00
5.31736791e-01 -1.82794213e-01 -7.74898767e-01 -1.39333948e-01
4.11790878e-01 7.11366475e-01 -5.99297941e-01 2.55331129e-01
-7.20173180e-01 2.32628703e-01 -3.71036232e-01 1.56840965e-01
-6.35465086e-01 -3.90201688e-01 8.43422771e-01 7.12684572e-01
1.71235134e-03 1.72545582e-01 -7.57552087e-01 -6.42336369e-01
5.49992472e-02 -8.67289186e-01 6.69530869e-01 -5.51155865e-01
-1.44264531e+00 1.05255425e-01 4.56854880e-01 -1.20556641e+00
-3.42339605e-01 -5.41208088e-01 -7.27867961e-01 1.41394269e+00
-1.34145474e+00 -1.02170289e+00 2.75935829e-01 2.83869833e-01
-2.31002152e-01 -2.51477212e-01 7.23535061e-01 1.10397391e-01
-1.44832641e-01 3.21251541e-01 -1.91452622e-01 -4.70927298e-01
2.64137477e-01 -1.40393066e+00 5.04212916e-01 4.43304271e-01
-5.66786408e-01 6.47991121e-01 4.64074552e-01 -1.21239114e+00
-7.93860972e-01 -5.83434939e-01 9.22171772e-01 -9.81834292e-01
8.45611811e-01 -2.36321032e-01 -5.25864482e-01 6.82769120e-01
3.47922072e-02 -8.68838787e-01 8.95216703e-01 2.09228411e-01
-8.24529454e-02 6.82228506e-02 -1.83096385e+00 3.42417389e-01
8.53036225e-01 -6.38212264e-01 -7.17409849e-01 6.96333423e-02
1.21110034e+00 4.01040167e-01 -1.49370253e+00 3.55952352e-01
5.38475156e-01 -1.31816494e+00 1.18448639e+00 -4.22786415e-01
1.32980481e-01 1.16538666e-01 -2.78237849e-01 -6.75332487e-01
-2.37283453e-01 -3.83800536e-01 3.81563872e-01 1.63116837e+00
7.16231704e-01 -1.22611034e+00 2.88860828e-01 1.19688129e+00
1.85763925e-01 -7.71695316e-01 -1.20550382e+00 -7.51174569e-01
3.19047272e-01 -8.04854155e-01 1.00654006e+00 9.24304962e-01
2.40926087e-01 -4.12881643e-01 1.17854722e-01 3.88598554e-02
7.51656234e-01 -2.37775251e-01 1.32255867e-01 -1.56175351e+00
-2.82926410e-01 -1.85142383e-01 -4.38065112e-01 4.28820938e-01
-5.06199636e-02 -4.13906574e-01 -6.60676360e-01 -1.42077398e+00
-8.43984708e-02 -4.99424160e-01 -6.86136961e-01 1.46333929e-02
8.33441615e-02 -4.46718037e-01 4.05886263e-01 8.38095844e-02
5.81106186e-01 9.63639244e-02 3.16974282e-01 3.97319943e-01
-8.33405405e-02 6.23426676e-01 -1.05261779e+00 6.03884757e-01
7.51342356e-01 -3.68701577e-01 -7.01741993e-01 3.88824344e-01
7.44864941e-01 6.10357642e-01 7.16346264e-01 -7.02437103e-01
-5.24226248e-01 -6.19521022e-01 -2.30915502e-01 -3.82955819e-01
-4.21813726e-02 -1.30953455e+00 7.62421429e-01 5.76654971e-01
-3.40745956e-01 2.52974451e-01 3.62046659e-02 9.43734944e-01
3.53916027e-02 -4.10629183e-01 2.02101782e-01 -7.86362439e-02
2.97889799e-01 -1.84592336e-01 -6.20744169e-01 -2.03692034e-01
1.49078190e+00 -3.77168804e-01 -9.32642519e-02 -3.61579031e-01
-7.69059777e-01 -2.03977063e-01 6.71556890e-01 3.89212698e-01
-4.61358577e-02 -1.36236787e+00 -4.21630532e-01 -3.03618073e-01
-1.78017430e-02 -6.33547843e-01 2.14607656e-01 4.51876163e-01
-2.28036284e-01 4.78886962e-01 -3.57175231e-01 4.12573248e-01
-7.51374841e-01 9.02581632e-01 3.16945225e-01 -3.09091002e-01
-3.86540085e-01 2.73738354e-01 2.18252346e-01 -1.94805995e-01
-2.86447871e-02 -6.57982171e-01 -3.19853723e-01 4.16246772e-01
1.24657132e-01 7.12339044e-01 -8.67835060e-02 -4.09732968e-01
-3.80425692e-01 1.70662552e-01 3.96119952e-01 -6.20763004e-01
1.35233545e+00 -2.12738350e-01 -5.98663352e-02 6.43376112e-01
7.25031435e-01 -8.66907761e-02 -1.37249708e+00 5.68892300e-01
4.81575608e-01 -1.47069305e-01 -3.32448781e-01 -1.12064612e+00
-6.52610600e-01 3.12756866e-01 7.01042652e-01 9.00094509e-01
1.31268454e+00 -7.10752383e-02 2.88715750e-01 -4.97014701e-01
5.40009916e-01 -1.22742951e+00 -5.40753424e-01 -3.19190353e-01
1.20388544e+00 -7.46455193e-01 -1.02868369e-02 -7.23192751e-01
-7.93964744e-01 3.71779710e-01 -1.39411300e-01 -3.90846550e-01
1.39182627e+00 3.03493172e-01 -2.63963807e-02 -2.54656166e-01
-5.73456347e-01 2.20870432e-02 8.31040815e-02 8.86814713e-01
4.08057690e-01 4.10396516e-01 -9.79516327e-01 1.36075151e+00
-1.75989702e-01 2.18192220e-01 5.68520606e-01 9.56203759e-01
-8.10614750e-02 -1.20840812e+00 -4.70159858e-01 5.73273242e-01
-6.93097770e-01 1.08298808e-01 -4.17985737e-01 9.17421699e-01
3.97830039e-01 1.16258991e+00 -1.06762022e-01 -3.33976835e-01
9.75377500e-01 2.35529348e-01 -1.32237732e-01 -4.42088664e-01
-1.04462814e+00 -3.98232669e-01 4.24665064e-01 -3.20522964e-01
-2.72500753e-01 -1.24191773e+00 -7.87283480e-01 -1.25692800e-01
-3.79400730e-01 -6.36332929e-02 7.66499996e-01 9.95398879e-01
2.79945374e-01 5.37436664e-01 5.59874713e-01 -4.39164460e-01
-9.80819225e-01 -7.60508537e-01 -1.01646781e+00 1.86524168e-01
6.33300021e-02 -6.26550019e-01 -9.38636065e-01 -2.59252399e-01] | [5.8678178787231445, 4.422704219818115] |
647d46fe-25dd-4754-8bb8-45cc27210aa3 | dialogxl-all-in-one-xlnet-for-multi-party | 2012.08695 | null | https://arxiv.org/abs/2012.08695v1 | https://arxiv.org/pdf/2012.08695v1.pdf | DialogXL: All-in-One XLNet for Multi-Party Conversation Emotion Recognition | This paper presents our pioneering effort for emotion recognition in conversation (ERC) with pre-trained language models. Unlike regular documents, conversational utterances appear alternately from different parties and are usually organized as hierarchical structures in previous work. Such structures are not conducive to the application of pre-trained language models such as XLNet. To address this issue, we propose an all-in-one XLNet model, namely DialogXL, with enhanced memory to store longer historical context and dialog-aware self-attention to deal with the multi-party structures. Specifically, we first modify the recurrence mechanism of XLNet from segment-level to utterance-level in order to better model the conversational data. Second, we introduce dialog-aware self-attention in replacement of the vanilla self-attention in XLNet to capture useful intra- and inter-speaker dependencies. Extensive experiments are conducted on four ERC benchmarks with mainstream models presented for comparison. The experimental results show that the proposed model outperforms the baselines on all the datasets. Several other experiments such as ablation study and error analysis are also conducted and the results confirm the role of the critical modules of DialogXL. | ['Zhixian Xie', 'Xiaojun Quan', 'Junqing Chen', 'Weizhou Shen'] | 2020-12-16 | null | null | null | null | ['emotion-recognition-in-conversation'] | ['natural-language-processing'] | [-7.21086934e-02 2.31459126e-01 5.27807102e-02 -8.01577806e-01
-6.09883010e-01 -3.19310218e-01 5.97683012e-01 1.31860882e-01
-5.18664479e-01 8.96425664e-01 8.74851823e-01 -2.61795104e-01
4.20405686e-01 -2.51385093e-01 -1.28732547e-01 -5.08327603e-01
-5.03993407e-02 3.44874233e-01 -2.36924127e-01 -5.17086864e-01
2.28594244e-01 1.79991990e-01 -1.15737653e+00 5.57144284e-01
9.43096876e-01 8.46317172e-01 2.07182139e-01 6.74317658e-01
-8.10255647e-01 1.09081662e+00 -8.16046596e-01 -5.05409956e-01
-2.32176408e-01 -6.05618775e-01 -1.02447271e+00 2.04359993e-01
-1.42759740e-01 -2.35046223e-01 -1.57915458e-01 7.23814368e-01
7.49305725e-01 5.24409950e-01 3.97850245e-01 -1.06932402e+00
-6.27611637e-01 1.02032483e+00 -4.48124766e-01 1.69375706e-02
4.36546803e-01 2.22644024e-02 1.10745037e+00 -1.02764571e+00
4.40403908e-01 1.70265651e+00 5.29345453e-01 7.43218005e-01
-7.28818715e-01 -5.42285442e-01 6.88946366e-01 2.64968276e-01
-9.51929510e-01 -6.14572823e-01 1.04317570e+00 -1.55807901e-02
1.39956427e+00 3.32228631e-01 2.95141250e-01 1.50771832e+00
1.82767987e-01 1.13514531e+00 9.54701543e-01 -6.62438631e-01
7.87108466e-02 3.77357095e-01 6.22826517e-01 4.62068558e-01
-5.78751564e-01 -2.97142833e-01 -6.84451759e-01 -2.90744036e-01
2.36545891e-01 -1.25184596e-01 -2.70760596e-01 1.95255294e-01
-7.43629158e-01 1.12586033e+00 1.09037951e-01 6.44068003e-01
-2.96728551e-01 -2.33015954e-01 9.62961495e-01 4.98441637e-01
7.21468568e-01 3.82032305e-01 -6.21940613e-01 -2.86157191e-01
-5.91758907e-01 -1.35870248e-01 1.04788744e+00 1.15639794e+00
4.71404284e-01 1.22925110e-01 -3.66026610e-01 1.24769950e+00
2.26899266e-01 -1.85936585e-01 9.68964517e-01 -7.27869093e-01
6.02024019e-01 7.87102997e-01 -2.51782715e-01 -6.40139282e-01
-3.48280430e-01 -3.12651098e-01 -1.03983545e+00 -3.88806880e-01
-1.65372521e-01 -4.79987621e-01 -6.70906663e-01 1.77249229e+00
7.61450231e-02 6.80779740e-02 5.77985168e-01 5.57386339e-01
1.22251499e+00 8.96716952e-01 3.47369373e-01 -5.33802271e-01
1.35682535e+00 -1.60460162e+00 -1.25762677e+00 -3.61526817e-01
7.03600466e-01 -8.39762986e-01 1.34841323e+00 1.38294384e-01
-1.03576088e+00 -5.98510087e-01 -8.23027670e-01 -1.75659314e-01
-4.57236946e-01 1.32528961e-01 4.75215584e-01 4.82867151e-01
-9.57914054e-01 1.49835095e-01 -5.24195850e-01 -4.59550858e-01
-1.34343967e-01 1.20003201e-01 -7.37857744e-02 4.02627170e-01
-1.62793672e+00 1.04470623e+00 4.23411101e-01 5.43618739e-01
-4.92191434e-01 -3.31081986e-01 -9.69142318e-01 1.90146476e-01
4.42655653e-01 -1.68810725e-01 1.49532151e+00 -1.07733643e+00
-1.97346616e+00 6.45300090e-01 -4.64648992e-01 -4.78782773e-01
2.68151581e-01 -5.67124784e-01 -5.14329076e-01 -1.51233718e-01
-2.43738562e-01 5.08888781e-01 4.48883384e-01 -1.37630820e+00
-5.78195989e-01 -7.60510191e-03 1.34202808e-01 5.82143366e-01
-3.52613539e-01 3.35222095e-01 -6.05993629e-01 -5.92830062e-01
-1.58489242e-01 -8.40917349e-01 -3.05187166e-01 -8.16729009e-01
-6.28637254e-01 -6.60348535e-01 1.01484931e+00 -6.25576556e-01
1.57245719e+00 -2.15060544e+00 8.02314878e-02 -1.86272308e-01
-1.12508073e-01 1.95632219e-01 -1.49815544e-01 8.29896629e-01
-1.47905961e-01 1.39740258e-01 -1.81730643e-01 -1.00642312e+00
6.64174184e-02 3.36716086e-01 -4.48340356e-01 1.37049947e-02
2.22375154e-01 9.63673890e-01 -4.99809325e-01 -5.92371047e-01
2.76411533e-01 3.23519081e-01 -3.97590011e-01 6.79132760e-01
-2.51255572e-01 6.14157677e-01 -3.40185106e-01 5.24595797e-01
3.85871887e-01 -2.25945991e-02 2.57962644e-01 -5.43405116e-02
-1.47591561e-01 6.69828773e-01 -7.80513525e-01 1.85314548e+00
-9.02800322e-01 5.23437023e-01 3.97025138e-01 -8.58555377e-01
1.01141715e+00 7.66910553e-01 1.46345571e-01 -4.78930265e-01
3.97697002e-01 -2.39857256e-01 1.06449015e-01 -4.92179126e-01
8.13244343e-01 -2.43370116e-01 -4.99422759e-01 4.88811851e-01
2.09294856e-01 8.32671002e-02 -1.62541792e-02 3.99023920e-01
7.55030692e-01 -2.64793545e-01 4.35968369e-01 -2.10279793e-01
9.91710067e-01 -3.35360259e-01 7.18643785e-01 6.17207944e-01
-3.43680739e-01 2.97353983e-01 5.06474733e-01 -1.20417021e-01
-4.18459415e-01 -5.06315231e-01 8.79046246e-02 1.66359472e+00
8.19376409e-02 -5.94935834e-01 -6.75595939e-01 -8.77306163e-01
-4.59582269e-01 1.06470764e+00 -4.80074048e-01 -8.02705735e-02
-5.70071876e-01 -7.08959699e-01 5.94218731e-01 5.46353042e-01
8.20104837e-01 -1.83002973e+00 -3.60722363e-01 5.57513237e-01
-5.59030116e-01 -1.08648419e+00 -6.22574627e-01 5.08315086e-01
-4.97283727e-01 -6.89169765e-01 -3.85223418e-01 -9.45280015e-01
2.60937631e-01 -1.89801559e-01 1.21510184e+00 1.16993926e-01
9.29708499e-03 3.74173135e-01 -6.78769708e-01 -4.67809826e-01
-4.29719776e-01 3.61416787e-01 -1.43534541e-01 -4.79703024e-03
5.90001345e-01 -5.24026394e-01 -3.44985992e-01 8.72228220e-02
-6.49172246e-01 -5.62256873e-02 5.59527695e-01 1.11922622e+00
-2.91357469e-02 -4.01586257e-02 1.04826355e+00 -1.22101545e+00
1.30080473e+00 -3.39948744e-01 -7.30093196e-02 4.71799940e-01
-2.37252027e-01 -2.96144783e-02 6.34631276e-01 -4.67608571e-01
-1.67787516e+00 -1.89150959e-01 -4.69692439e-01 -1.90518722e-01
-2.30761960e-01 6.17345989e-01 -6.19864702e-01 5.27725279e-01
1.12880155e-01 1.88674718e-01 -2.54599690e-01 -5.49000859e-01
4.56905991e-01 1.01536095e+00 4.40052748e-01 -7.80231953e-01
8.87586772e-02 -1.17338695e-01 -9.66510832e-01 -9.20258343e-01
-9.09017026e-01 -5.62664807e-01 -5.84196031e-01 -1.76070228e-01
1.01826012e+00 -8.78875196e-01 -8.17389429e-01 3.75394017e-01
-1.49405372e+00 -3.70399088e-01 -1.43495768e-01 3.58317226e-01
-2.41017759e-01 3.66066307e-01 -1.21209681e+00 -1.24355042e+00
-7.23914683e-01 -1.14187646e+00 1.03978157e+00 3.25924635e-01
-4.22298968e-01 -1.24336958e+00 5.35836294e-02 2.25979403e-01
5.14520288e-01 -2.02059418e-01 9.82566893e-01 -1.08226180e+00
-8.04814398e-02 1.35421142e-01 1.21116377e-01 5.69913983e-01
1.76039740e-01 -2.33930707e-01 -1.18662238e+00 -1.96323425e-01
4.37701732e-01 -7.10116506e-01 6.95977271e-01 -4.15618531e-02
1.20191836e+00 -3.74692172e-01 -1.51492953e-01 7.57284686e-02
1.07276750e+00 5.62074959e-01 5.87679446e-01 3.75597998e-02
4.56464678e-01 1.00708413e+00 5.40851772e-01 3.64780426e-01
5.07048130e-01 4.15516973e-01 4.54703830e-02 -2.82335490e-01
3.16893160e-01 -2.99320221e-02 4.67323422e-01 1.45944083e+00
3.49342734e-01 -6.50036454e-01 -6.32907331e-01 5.08260131e-01
-2.00907016e+00 -7.86493957e-01 4.74747159e-02 1.67841530e+00
1.01657403e+00 1.95489585e-01 -1.98439196e-01 -3.47966731e-01
7.36329675e-01 3.91721636e-01 -4.69746619e-01 -8.45001757e-01
-2.58408099e-01 7.22301602e-02 -1.59624159e-01 8.64413977e-01
-9.84160960e-01 1.23356831e+00 5.97726107e+00 6.40090942e-01
-1.08221233e+00 2.12482959e-01 8.90148997e-01 1.61158834e-02
-1.49375558e-01 -1.24409506e-02 -9.70084071e-01 3.57426286e-01
1.07239997e+00 -1.62744388e-01 8.22282881e-02 8.89330208e-01
1.08993128e-01 -2.84316670e-02 -1.19152391e+00 9.01528895e-01
2.39949241e-01 -9.41500723e-01 -1.90562323e-01 -1.56556010e-01
4.53192890e-01 -2.27597013e-01 -3.22595835e-01 1.06728470e+00
5.77542424e-01 -9.28111970e-01 3.51601064e-01 2.77002007e-01
2.62949675e-01 -9.25729811e-01 1.06738102e+00 4.18749273e-01
-1.21345615e+00 1.26377538e-01 -4.11601305e-01 3.88499685e-02
5.00949979e-01 2.23907247e-01 -8.04336727e-01 7.12193191e-01
5.77461064e-01 4.82437521e-01 -2.87221104e-01 2.82795608e-01
-1.81429744e-01 7.29035974e-01 -7.46671343e-03 -2.12427557e-01
5.20519376e-01 -3.07866663e-01 4.08173382e-01 1.60854805e+00
1.91407073e-02 3.09602797e-01 2.86461204e-01 5.45883715e-01
-2.53210336e-01 5.15438139e-01 -4.32797372e-01 5.99295571e-02
5.24488211e-01 1.31721401e+00 -3.19464028e-01 -6.03070617e-01
-7.15470672e-01 1.17548490e+00 4.77932304e-01 5.17348766e-01
-8.04476321e-01 -3.43880951e-01 5.25874436e-01 -3.75586361e-01
1.16646014e-01 -8.28171745e-02 5.42955138e-02 -1.07953846e+00
-1.02138601e-01 -1.10222995e+00 4.34494406e-01 -4.95810956e-01
-1.70044541e+00 1.24594963e+00 -1.73835352e-01 -7.22598732e-01
-5.06861150e-01 -1.92061827e-01 -1.18480885e+00 1.00269508e+00
-1.39159608e+00 -9.92482603e-01 -1.39165834e-01 6.62012756e-01
1.27415752e+00 -3.15518439e-01 1.09985924e+00 1.89225525e-01
-8.54315400e-01 6.24320328e-01 -3.33824120e-02 1.41336381e-01
9.91247416e-01 -1.33684969e+00 1.82213679e-01 6.13165259e-01
-1.31911933e-01 9.15506303e-01 5.64538836e-01 -5.71023703e-01
-9.40113008e-01 -1.04943419e+00 1.19266784e+00 -2.19565541e-01
5.85213125e-01 -8.40358257e-01 -1.22746181e+00 9.33835328e-01
1.12838566e+00 -6.13830030e-01 9.69255328e-01 7.08647668e-01
4.35556881e-02 1.14257812e-01 -8.23116362e-01 6.86709285e-01
7.92029858e-01 -6.82493925e-01 -9.85948801e-01 1.84160426e-01
1.03689456e+00 -2.57814616e-01 -5.42169333e-01 4.64475095e-01
2.12579012e-01 -8.68892848e-01 5.17403245e-01 -6.00575566e-01
3.04174036e-01 2.76029885e-01 -1.55060843e-01 -1.34942961e+00
1.26230558e-02 -9.05338883e-01 5.45397885e-02 1.82342505e+00
4.87755537e-01 -4.64154840e-01 6.48587227e-01 7.83213258e-01
-5.12019098e-01 -7.50323951e-01 -7.96167433e-01 -4.05065745e-01
3.20678540e-02 -3.25599611e-01 4.36850220e-01 1.05182326e+00
3.39500755e-01 1.17821217e+00 -6.94324434e-01 -1.22806109e-01
-1.55881541e-02 1.99561805e-01 7.44878232e-01 -8.84765863e-01
-2.44192377e-01 -2.88975865e-01 2.79205680e-01 -1.34518826e+00
6.98895752e-01 -6.54741347e-01 4.70587015e-01 -1.32979500e+00
1.83109920e-02 -2.28788152e-01 -2.29887053e-01 3.22700411e-01
-3.81875694e-01 -3.65566134e-01 1.76326260e-01 -5.89119308e-02
-7.61382580e-01 1.23582089e+00 8.50091338e-01 -8.96694511e-03
-6.31929934e-01 -4.21146005e-02 -8.78328443e-01 7.63312995e-01
8.89693677e-01 -1.65010318e-01 -6.57316804e-01 -1.60378575e-01
-4.59169149e-01 2.71743149e-01 -2.22399145e-01 -5.70279419e-01
3.98523092e-01 9.36998948e-02 -4.12037317e-03 -9.97307241e-01
5.29100358e-01 -7.45274007e-01 -3.59961659e-01 -1.07033392e-02
-7.79576421e-01 2.42290720e-01 3.76085907e-01 4.74051505e-01
-6.22567236e-01 -3.43771279e-01 5.23176253e-01 -2.44327396e-01
-6.17560327e-01 -6.98601529e-02 -6.76224530e-01 9.68226716e-02
8.19854200e-01 2.25634873e-01 -1.64161652e-01 -8.25842261e-01
-1.00773215e+00 7.08615959e-01 -1.50915533e-01 6.66581035e-01
4.75081533e-01 -1.12110651e+00 -6.12679839e-01 4.71330434e-02
1.36835590e-01 -7.73740709e-02 4.70006287e-01 4.19939458e-01
9.33342502e-02 5.71539462e-01 6.12820163e-02 -3.28065574e-01
-1.46088576e+00 4.19766635e-01 3.49872857e-01 -8.30796719e-01
-4.51464325e-01 9.18051243e-01 6.10859096e-01 -8.03246319e-01
8.39382887e-01 -4.23527688e-01 -3.53921264e-01 9.68915671e-02
3.55171174e-01 -1.27286762e-01 -4.50415686e-02 -5.15790522e-01
-4.14357901e-01 -4.25592028e-02 -5.99518061e-01 -3.09411645e-01
1.21747077e+00 -5.96298635e-01 -2.09721804e-01 8.57577801e-01
1.11638737e+00 1.11372076e-01 -1.10350609e+00 -4.73257750e-01
2.79396892e-01 1.08598612e-01 -1.40246943e-01 -9.31396067e-01
-7.93802917e-01 9.36091304e-01 2.68622458e-01 1.99408650e-01
1.09718740e+00 -2.18053795e-02 9.05428350e-01 5.32303452e-01
-7.20023215e-02 -1.34504402e+00 3.11127782e-01 9.90537167e-01
1.34629381e+00 -1.26979172e+00 -3.97501528e-01 -3.99978608e-01
-1.18930209e+00 8.99629474e-01 1.09144151e+00 1.10245138e-01
6.82886004e-01 3.70288044e-01 4.38955098e-01 -1.76038638e-01
-1.30896151e+00 -1.14129409e-01 5.17861247e-02 4.77450900e-02
9.71549809e-01 -2.99307346e-01 -5.42087853e-01 9.98096526e-01
-1.99094713e-01 -4.40723956e-01 2.44747594e-01 1.01370454e+00
-2.91120380e-01 -1.39188695e+00 7.19943196e-02 1.42194793e-01
-5.29903710e-01 -4.38769966e-01 -7.90865481e-01 8.77046406e-01
-2.01413304e-01 1.27385426e+00 4.82315570e-02 -2.51030713e-01
3.73364180e-01 6.11607969e-01 -1.92041576e-01 -7.76773870e-01
-1.11385131e+00 4.17772055e-01 4.52027977e-01 -3.31346571e-01
-5.32046258e-01 -3.50729465e-01 -1.26647377e+00 -7.73732811e-02
-4.87276822e-01 6.16185009e-01 4.68515694e-01 9.59938288e-01
3.52671415e-01 7.77817428e-01 6.81899607e-01 -5.20948172e-01
-4.80406076e-01 -1.59959352e+00 -4.31433827e-01 3.78176272e-01
2.17675045e-01 -3.59144270e-01 -4.45878983e-01 -6.84856698e-02] | [12.95085620880127, 6.373098373413086] |
d36aec68-17ac-4b17-8b12-c734f08a667f | a-study-of-graph-based-approaches-for-semi | 2104.08153 | null | https://arxiv.org/abs/2104.08153v2 | https://arxiv.org/pdf/2104.08153v2.pdf | An Empirical Study of Graph-Based Approaches for Semi-Supervised Time Series Classification | Time series data play an important role in many applications and their analysis reveals crucial information for understanding the underlying processes. Among the many time series learning tasks of great importance, we here focus on semi-supervised learning based on a graph representation of the data. Two main aspects are involved in this task. A suitable distance measure to evaluate the similarities between time series, and a learning method to make predictions based on these distances. However, the relationship between the two aspects has never been studied systematically in the context of graph-based learning. We describe four different distance measures, including (Soft) DTW and MPDist, a distance measure based on the Matrix Profile, as well as four successful semi-supervised learning methods, including the graph Allen--Cahn method and a Graph Convolutional Neural Network. We then compare the performance of the algorithms on binary classification data sets. In our findings we compare the chosen graph-based methods using all distance measures and observe that the results vary strongly with respect to the accuracy. As predicted by the ``no free lunch'' theorem, no clear best combination to employ in all cases is found. Our study provides a reproducible framework for future work in the direction of semi-supervised learning for time series with a focus on graph representations. | ['Martin Stoll', 'Lucile Peroche', 'Miriam Gondos', 'Dominik Alfke'] | 2021-04-16 | null | null | null | null | ['semi-supervised-time-series-classification'] | ['time-series'] | [ 1.47849396e-01 1.23897217e-01 -1.34708181e-01 -2.19101369e-01
-1.87986001e-01 -6.01975024e-01 8.80564392e-01 9.01005387e-01
-3.77731442e-01 3.80677462e-01 -2.18754280e-02 -5.30491173e-01
-7.87020922e-01 -9.19325411e-01 -2.85037637e-01 -9.01057363e-01
-8.52527320e-01 4.46586847e-01 2.78033793e-01 -4.80117261e-01
2.09962279e-01 5.46796381e-01 -1.45081925e+00 3.37577164e-02
6.07633531e-01 1.13443244e+00 -1.80655882e-01 5.19537270e-01
-1.23322345e-01 8.26342404e-01 -4.08918977e-01 -3.26176465e-01
3.61471653e-01 -6.68460131e-01 -7.23844230e-01 -2.00950112e-02
-1.39790233e-02 6.20909333e-01 -3.34696770e-01 8.43727291e-01
2.74477810e-01 1.58902779e-01 9.63945866e-01 -1.30970883e+00
-1.63927421e-01 7.44461119e-01 -2.50919789e-01 4.30341601e-01
3.69666904e-01 -9.15811360e-02 1.16352153e+00 -2.62625277e-01
7.46363878e-01 7.83075690e-01 8.57825518e-01 -1.57777697e-01
-1.41976964e+00 -3.10694188e-01 -1.89689398e-01 5.15694082e-01
-1.17040122e+00 -5.61081618e-02 1.21275210e+00 -8.08139324e-01
7.44501114e-01 2.21274465e-01 7.05881894e-01 8.44749689e-01
3.68753046e-01 3.89222324e-01 1.47501624e+00 -6.02589607e-01
4.32202816e-01 -1.33103773e-01 2.70000249e-01 7.40570426e-01
2.17065528e-01 2.48223450e-02 -1.18628107e-01 -3.11533511e-02
4.38874573e-01 6.33038729e-02 -2.44488701e-01 -7.79208004e-01
-1.23363376e+00 1.01343203e+00 5.42499006e-01 8.71007502e-01
-1.74983799e-01 -3.88680696e-02 7.56213427e-01 8.72089088e-01
7.08430409e-01 5.62427282e-01 -2.59008765e-01 5.51628470e-02
-9.28922117e-01 -1.43261747e-02 9.60646212e-01 2.04105496e-01
6.82867229e-01 -2.37829499e-02 1.19108237e-01 5.12878299e-01
6.12376630e-02 7.09645301e-02 5.28389275e-01 -3.45891744e-01
4.17457700e-01 7.36161411e-01 -4.55672264e-01 -1.43053377e+00
-9.35995519e-01 -5.62131763e-01 -1.11444187e+00 2.53129810e-01
8.10203671e-01 -4.01748940e-02 -3.45028102e-01 1.38112462e+00
2.46599138e-01 1.26225829e-01 -2.58695066e-01 6.28671110e-01
6.96355343e-01 4.22718197e-01 -2.09688708e-01 -4.98448282e-01
9.40178156e-01 -5.21192908e-01 -5.86436212e-01 3.27230901e-01
8.54554057e-01 -5.86004674e-01 7.53745139e-01 4.63147491e-01
-6.40093625e-01 -5.03529608e-01 -1.18519115e+00 4.42503989e-01
-8.37094367e-01 -2.53601577e-02 6.85668170e-01 5.43594658e-01
-1.06305289e+00 1.38360071e+00 -8.54448617e-01 -7.10125387e-01
1.30112499e-01 4.25749719e-01 -4.96787876e-01 2.24018916e-01
-1.11055934e+00 9.14578438e-01 3.76590282e-01 -7.54750594e-02
-2.67951041e-01 -2.32910991e-01 -8.48821878e-01 3.73862833e-02
1.42816111e-01 -2.15400815e-01 6.27592087e-01 -7.93887079e-01
-1.27310574e+00 9.95332956e-01 4.40328598e-01 -9.29883182e-01
7.54775345e-01 5.47265589e-01 -5.39995670e-01 1.12362944e-01
-1.81141093e-01 5.75008467e-02 8.00430417e-01 -7.47327507e-01
-1.53695151e-01 -3.78237873e-01 -1.37782440e-01 -3.61591399e-01
-2.82383293e-01 -2.27672726e-01 2.25726664e-01 -7.84879982e-01
1.87429607e-01 -1.03731728e+00 -2.54890710e-01 -1.62782744e-01
-1.67366818e-01 -4.16072160e-01 7.55409479e-01 -4.08177823e-01
1.34357917e+00 -1.96252215e+00 3.18232685e-01 4.78099138e-01
4.59927619e-01 2.27038041e-01 -4.16768715e-02 1.18667686e+00
-5.41522443e-01 -1.09827332e-01 -3.79674643e-01 -1.36358827e-01
1.14071429e-01 -7.36832805e-03 -2.55505741e-01 8.62072170e-01
5.18679693e-02 6.70882761e-01 -1.00606322e+00 -2.77966678e-01
3.80768418e-01 4.14509714e-01 -1.74010489e-02 1.40873998e-01
4.75266343e-03 3.50273281e-01 -1.41711712e-01 3.65451984e-02
2.38994226e-01 -3.34954470e-01 3.53740662e-01 -1.97701275e-01
-1.32127315e-01 1.86894819e-01 -9.72632527e-01 1.40056884e+00
-1.50968239e-01 8.36963236e-01 -4.70527381e-01 -1.74023342e+00
1.19288397e+00 2.36184970e-01 8.98338020e-01 -6.26980722e-01
2.42852122e-01 3.06533158e-01 3.37143004e-01 -3.88994992e-01
5.57972351e-03 -6.92883506e-02 8.86169896e-02 6.65409446e-01
2.13872701e-01 -2.08871462e-03 5.95614314e-01 2.13328507e-02
1.40280795e+00 -1.65169928e-02 6.27984941e-01 -4.59212184e-01
5.98258376e-01 4.12144475e-02 -4.61116359e-02 3.01050663e-01
-1.34726420e-01 4.77775007e-01 1.02578557e+00 -7.03737617e-01
-9.59282458e-01 -7.04073429e-01 -1.10254087e-01 8.63809407e-01
-2.05783620e-01 -7.45479703e-01 -5.01947761e-01 -7.58824885e-01
5.17537072e-02 3.77672523e-01 -8.59722435e-01 -2.46112943e-01
-3.86150450e-01 -7.58904815e-01 2.88649231e-01 2.16115460e-01
1.15694501e-01 -9.41096544e-01 -4.93692726e-01 6.79016113e-02
2.03626975e-01 -1.00399888e+00 -7.15422407e-02 6.72436357e-01
-1.16888118e+00 -1.40592420e+00 -5.32374620e-01 -6.45980895e-01
3.46723884e-01 2.06201404e-01 1.08778191e+00 2.27729734e-02
-1.92778453e-01 3.07099462e-01 -5.76481223e-01 -2.09839106e-01
-6.34449184e-01 1.45277947e-01 -1.16508782e-01 2.45813712e-01
2.69903988e-01 -9.40999210e-01 -4.48956668e-01 2.10797250e-01
-8.72778416e-01 -2.61334777e-01 4.56247896e-01 5.26718676e-01
3.54951620e-01 4.37071979e-01 2.65970886e-01 -9.22803581e-01
7.88321793e-01 -5.32456040e-01 -5.54855168e-01 1.29416078e-01
-7.83075392e-01 3.27903390e-01 9.44866121e-01 -3.52571040e-01
-5.06881773e-02 3.95426340e-02 7.71124288e-02 -4.55044627e-01
-2.22345646e-02 7.81752288e-01 3.47360402e-01 -3.47959459e-01
6.42464459e-01 1.68926612e-01 1.93948567e-01 -3.57250512e-01
3.08416039e-01 3.02257031e-01 1.02757894e-01 -2.38826960e-01
8.00263166e-01 5.28408229e-01 5.56214333e-01 -1.06120813e+00
-6.16436720e-01 -6.20236874e-01 -1.01881051e+00 -3.55608940e-01
7.95621276e-01 -3.57386738e-01 -6.71708107e-01 2.27114826e-01
-8.05036604e-01 -4.29540485e-01 -3.58108670e-01 5.72204709e-01
-7.45340049e-01 5.66858292e-01 -5.00097036e-01 -7.54625499e-01
-2.11049989e-01 -7.83013582e-01 7.80722439e-01 -3.06179523e-01
-2.68451095e-01 -1.51861930e+00 5.34309745e-01 -3.27435508e-02
3.37210506e-01 7.72702515e-01 1.23199153e+00 -1.07980144e+00
-9.27662477e-02 -4.08835679e-01 -6.11913279e-02 1.70692652e-01
2.16791973e-01 4.18290757e-02 -6.08213007e-01 -3.79766554e-01
-1.85007360e-02 2.81165987e-02 1.05326760e+00 3.07566285e-01
8.40093791e-01 -6.75746873e-02 -2.57022679e-01 4.00993139e-01
1.52655244e+00 1.07389286e-01 4.13754523e-01 2.49334782e-01
6.48828208e-01 7.75788426e-01 2.62366563e-01 4.67805326e-01
1.77747056e-01 6.61297023e-01 4.48695272e-01 7.91132823e-03
-4.78536859e-02 -2.04692796e-01 3.50538850e-01 1.10308003e+00
-3.95142674e-01 -1.08642951e-01 -1.02337098e+00 3.05714786e-01
-1.98458266e+00 -1.12615585e+00 -5.45683682e-01 2.46900797e+00
3.15172940e-01 4.91990089e-01 5.61006308e-01 8.77872646e-01
6.39406383e-01 5.35132349e-01 -1.10332547e-02 -4.34628844e-01
-1.17915817e-01 4.10093814e-01 3.65939111e-01 2.91683793e-01
-1.36647165e+00 2.92692095e-01 5.95203686e+00 5.51756084e-01
-1.42019832e+00 -1.25556886e-01 3.40997875e-01 3.87817144e-01
2.23654900e-02 -3.32046971e-02 -1.97005570e-01 3.62102687e-01
1.21421123e+00 -2.34736025e-01 4.20045137e-01 6.46268666e-01
2.78213382e-01 2.02517390e-01 -1.22992146e+00 9.90656912e-01
1.26714140e-01 -1.16964018e+00 -2.66295671e-01 1.32992283e-01
5.15718341e-01 7.21885264e-02 -8.50380212e-02 1.69784188e-01
-5.63555509e-02 -8.96362603e-01 4.53798592e-01 5.51508069e-01
3.18769097e-01 -6.11079335e-01 6.09358251e-01 1.87777981e-01
-1.47429216e+00 -1.43512767e-02 -1.55620441e-01 -3.50831330e-01
-1.56329408e-01 9.11199749e-01 -7.70625114e-01 6.95807993e-01
2.57490993e-01 1.13587475e+00 -7.26694822e-01 1.14548886e+00
-6.25080243e-02 6.94865286e-01 -2.41403058e-01 -2.36912593e-01
2.73255140e-01 -5.39282382e-01 5.01404285e-01 1.17540455e+00
3.33745599e-01 -3.96262437e-01 2.64412820e-01 6.05597138e-01
3.56206983e-01 3.85661960e-01 -9.42849278e-01 -3.54193062e-01
-9.31074247e-02 1.26513052e+00 -1.35426939e+00 -3.96831594e-02
-4.86188889e-01 6.28188968e-01 4.35284734e-01 1.30586270e-02
-5.28322518e-01 -4.39472467e-01 2.12357834e-01 4.61407542e-01
3.47514778e-01 -6.74038827e-01 -6.52756244e-02 -1.09682262e+00
-6.36575148e-02 -6.07812226e-01 7.65530467e-01 -4.87129390e-01
-1.54550171e+00 7.22847521e-01 3.89734469e-02 -1.55149174e+00
-4.29539412e-01 -8.46038043e-01 -6.53065145e-01 4.03150797e-01
-1.36225557e+00 -8.43089759e-01 -2.21895918e-01 6.50402665e-01
1.71062529e-01 -6.07203916e-02 7.32530951e-01 2.43950665e-01
-3.51259619e-01 2.27159802e-02 2.10607424e-01 2.21205503e-01
5.03218353e-01 -1.50837505e+00 2.72622287e-01 5.86044014e-01
5.78511477e-01 2.39225700e-01 8.40999842e-01 -3.70104849e-01
-1.48616815e+00 -8.84632885e-01 9.91209030e-01 -2.86283016e-01
1.20202780e+00 -4.06463921e-01 -1.05235529e+00 3.92925501e-01
1.52470857e-01 2.36565232e-01 6.81090593e-01 1.77551001e-01
-3.24825704e-01 -2.01316446e-01 -6.49936914e-01 3.10451269e-01
9.92143393e-01 -5.92583895e-01 -5.36502659e-01 6.29316986e-01
3.49467814e-01 1.83324412e-01 -1.05069900e+00 2.93075413e-01
3.62452477e-01 -1.34309363e+00 8.11369538e-01 -6.08383954e-01
3.95764679e-01 -7.25196078e-02 8.55617598e-02 -1.48876619e+00
-4.13548946e-01 -6.27954364e-01 -1.51712358e-01 9.59473729e-01
3.10980946e-01 -7.86335111e-01 6.60660207e-01 -1.47361726e-01
2.17093825e-01 -8.44693482e-01 -7.13917077e-01 -1.04011118e+00
2.84048263e-02 -4.79446858e-01 3.17002058e-01 1.27306986e+00
4.25655693e-01 5.30721247e-01 -1.47302002e-01 -2.42779762e-01
6.46343410e-01 4.44296092e-01 6.38203681e-01 -1.72065258e+00
-1.93424955e-01 -9.49165463e-01 -1.00845182e+00 -2.90333420e-01
1.46220773e-01 -1.27948880e+00 -4.50582147e-01 -1.37867093e+00
-2.50986993e-01 -3.50204885e-01 -4.98333037e-01 1.35033220e-01
2.61656642e-01 2.43420098e-02 5.53606600e-02 2.66037554e-01
-5.09202659e-01 4.14922327e-01 9.63929534e-01 -1.07608043e-01
-2.39220142e-01 2.19269991e-01 -7.92298838e-02 5.51047325e-01
7.62401640e-01 -4.09919351e-01 -4.88651454e-01 1.51649803e-01
3.31810802e-01 2.79295802e-01 2.96463847e-01 -1.25906003e+00
2.21451178e-01 1.44073039e-01 2.97442023e-02 -2.70981193e-01
-3.13295126e-02 -9.15661216e-01 1.95667967e-01 7.60095716e-01
-4.06722307e-01 4.80619937e-01 -2.26172447e-01 6.79983199e-01
-3.03222835e-01 -9.87577364e-02 6.59065664e-01 -8.27140361e-02
-6.37547553e-01 2.39521772e-01 -1.50332287e-01 -8.12352598e-02
9.86653328e-01 -1.69739664e-01 -5.17253652e-02 -5.75757205e-01
-9.04067814e-01 -1.02357768e-01 1.73090026e-01 4.03619409e-01
2.64737189e-01 -1.33497846e+00 -6.76495731e-01 1.66696683e-01
2.50106782e-01 -6.83776617e-01 -1.04685590e-01 1.28858519e+00
-5.52912831e-01 5.72266221e-01 -3.34872693e-01 -7.05097973e-01
-1.21298325e+00 1.02989233e+00 2.85213977e-01 -6.90669954e-01
-5.38982391e-01 1.93685383e-01 -2.80274987e-01 -1.89796656e-01
1.36913583e-01 -4.91685599e-01 -5.46198368e-01 5.78450382e-01
8.53098407e-02 4.82792616e-01 4.38560665e-01 -6.74303949e-01
-2.48871624e-01 7.57355571e-01 4.15906936e-01 1.75285399e-01
1.62201738e+00 5.72763346e-02 -2.84159690e-01 8.48596454e-01
1.47675991e+00 -2.12739408e-01 -6.71257138e-01 -2.26716697e-01
4.62332159e-01 -1.49964690e-01 -5.55091165e-03 -1.90027490e-01
-1.00677669e+00 9.56091642e-01 4.95260000e-01 1.29257381e+00
1.18334389e+00 -7.39609152e-02 3.40895295e-01 2.98729271e-01
4.53793436e-01 -7.35032201e-01 8.63729492e-02 3.48953664e-01
8.52855146e-01 -1.24546945e+00 2.37878174e-01 -4.39532220e-01
-3.74047339e-01 1.47486103e+00 -9.81447622e-02 -4.67212498e-01
1.04983974e+00 -1.05178215e-01 -1.88103944e-01 -4.54569846e-01
-6.74594998e-01 -5.99250853e-01 6.35087192e-01 6.11942172e-01
6.54593408e-01 6.59759045e-02 -7.96555817e-01 1.32341251e-01
-4.02231842e-01 -1.80006966e-01 3.80544752e-01 6.01035893e-01
-2.34154373e-01 -1.21427858e+00 -1.21355921e-01 5.06026268e-01
-2.21465558e-01 2.48215795e-01 -6.36504948e-01 8.87434304e-01
-9.61643010e-02 8.80636573e-01 -8.76110867e-02 -6.78393304e-01
2.33550012e-01 7.11698681e-02 5.62129498e-01 -4.42203462e-01
-7.30115592e-01 -1.57675698e-01 7.55031779e-02 -4.32384402e-01
-7.50754356e-01 -6.91999257e-01 -8.69465530e-01 -3.95612866e-01
-2.33218923e-01 2.07584694e-01 6.74880564e-01 1.01709807e+00
-6.40741957e-04 4.00383860e-01 8.81089330e-01 -8.27580810e-01
-3.39888543e-01 -7.50193059e-01 -8.38339865e-01 6.74593210e-01
1.36793315e-01 -5.82023740e-01 -5.32217085e-01 -1.91884249e-01] | [7.393445014953613, 3.6275107860565186] |
a2448a22-6e11-4b7f-9e2b-c51b6e9ea865 | an-intelligent-mechanism-for-monitoring-and | 2306.17187 | null | https://arxiv.org/abs/2306.17187v1 | https://arxiv.org/pdf/2306.17187v1.pdf | An Intelligent Mechanism for Monitoring and Detecting Intrusions in IoT Devices | The current amount of IoT devices and their limitations has come to serve as a motivation for malicious entities to take advantage of such devices and use them for their own gain. To protect against cyberattacks in IoT devices, Machine Learning techniques can be applied to Intrusion Detection Systems. Moreover, privacy related issues associated with centralized approaches can be mitigated through Federated Learning. This work proposes a Host-based Intrusion Detection Systems that leverages Federated Learning and Multi-Layer Perceptron neural networks to detected cyberattacks on IoT devices with high accuracy and enhancing data privacy protection. | ['Carlos Bento', 'Paulo Silva', 'Vitalina Holubenko'] | 2023-06-23 | null | null | null | null | ['intrusion-detection'] | ['miscellaneous'] | [-4.49777618e-02 2.31868222e-01 -6.47409201e-01 -5.01141429e-01
-1.86185017e-01 -8.35515976e-01 4.43144530e-01 3.22796226e-01
-3.89544755e-01 5.58248162e-01 -1.22419581e-01 -7.55253494e-01
-4.02125157e-03 -9.91433203e-01 -4.41726416e-01 -4.88235682e-01
2.06669830e-02 -6.21013790e-02 6.97618574e-02 1.16210088e-01
2.72662863e-02 8.28120530e-01 -1.21270394e+00 3.70377541e-01
4.83162522e-01 1.45680547e+00 -1.00423539e+00 5.31798184e-01
1.68245460e-03 6.84777915e-01 -5.55108488e-01 -6.96678758e-01
7.37502813e-01 2.34150290e-01 -4.66335863e-01 -4.96694893e-01
-1.23269148e-01 -5.17268896e-01 -4.33282286e-01 1.23889482e+00
1.43834874e-01 -9.27880332e-02 4.92139719e-02 -1.87295163e+00
-5.26757956e-01 5.11775315e-01 -1.35651395e-01 1.84704214e-02
3.96351039e-01 3.03371012e-01 4.89518851e-01 -2.35318635e-02
1.38638154e-01 9.99410331e-01 4.59733635e-01 6.60506248e-01
-8.75925004e-01 -1.10480320e+00 4.42607962e-02 -2.89132185e-02
-9.22421038e-01 -2.20132142e-01 8.90358210e-01 -1.13537244e-01
9.04506505e-01 5.48597515e-01 4.90056962e-01 1.22826517e+00
6.70550048e-01 4.69310164e-01 1.04172301e+00 -2.81639755e-01
7.17166841e-01 4.60450411e-01 4.42671627e-01 3.56954753e-01
1.15597522e+00 5.64715207e-01 1.22296788e-01 -6.06931865e-01
-7.21312016e-02 8.85136306e-01 3.20556313e-01 -2.85413265e-01
-4.57989365e-01 8.76476943e-01 5.89745820e-01 3.75689179e-01
-5.82218707e-01 -2.69980460e-01 9.95457113e-01 4.10501093e-01
1.00353226e-01 4.65970725e-01 -1.11446953e+00 1.36568666e-01
-2.00621039e-01 -3.19220304e-01 1.15680516e+00 5.82767606e-01
6.22392118e-01 1.03172563e-01 5.35391092e-01 -6.62368685e-02
6.22603536e-01 5.86664736e-01 4.36048925e-01 -5.98419070e-01
-9.55425389e-03 1.32758927e+00 -9.95154008e-02 -9.09499705e-01
-2.15262577e-01 1.18805282e-01 -7.61914372e-01 4.53159153e-01
-4.28946177e-03 -3.54081154e-01 -5.78486741e-01 1.27085280e+00
7.36250043e-01 1.74791932e-01 3.44339043e-01 3.73647362e-01
-2.61175279e-02 2.94802725e-01 4.73907799e-01 1.42655730e-01
1.38331008e+00 -3.58348221e-01 -6.83114707e-01 -3.43387462e-02
5.89628279e-01 -1.65229931e-01 4.73146260e-01 4.51466501e-01
-4.17746425e-01 2.68274378e-02 -1.30714345e+00 3.43004644e-01
-1.20760298e+00 -6.61620259e-01 1.01873159e+00 1.56318223e+00
-3.25192273e-01 4.67336118e-01 -1.07489109e+00 -2.60710597e-01
1.03049338e+00 1.07261705e+00 -2.89323896e-01 1.04323596e-01
-9.37546253e-01 7.55944312e-01 8.77221942e-01 -3.06729704e-01
-5.31116843e-01 -4.85899955e-01 -7.71387339e-01 -6.81481808e-02
1.88631341e-01 -5.02581596e-01 9.22636271e-01 -7.06865132e-01
-1.26448917e+00 5.87346792e-01 8.09846640e-01 -9.24235046e-01
7.12146461e-02 -1.23644091e-01 -1.29463613e+00 3.09706032e-02
-2.10028142e-01 -2.87188198e-02 9.43220079e-01 -7.62807846e-01
-7.12279856e-01 -7.85069227e-01 8.86966735e-02 -5.34267187e-01
-8.09134901e-01 2.18992189e-01 8.09176743e-01 -1.01397231e-01
-1.71736211e-01 -9.94158030e-01 -2.67395854e-01 -4.17552628e-02
-4.42563504e-01 2.67850682e-02 1.92658138e+00 -3.61707538e-01
8.63147438e-01 -1.99853039e+00 -7.81310380e-01 2.84225434e-01
2.57397443e-01 8.77519548e-01 2.90855408e-01 2.85992086e-01
1.48572261e-02 5.19757032e-01 3.88752013e-01 -7.67881721e-02
1.99445546e-01 2.92352766e-01 -3.39258999e-01 4.76293504e-01
-1.28155813e-01 9.09269810e-01 -5.12509942e-01 -2.46069014e-01
6.88576579e-01 4.31725264e-01 -1.40503421e-01 1.29774630e-01
-2.98876613e-01 4.73103791e-01 -1.10536945e+00 1.10433519e+00
8.75111997e-01 -2.32724622e-01 3.11257899e-01 -1.05422772e-01
-2.56852228e-02 1.54634953e-01 -6.18875384e-01 1.01937759e+00
-2.77401656e-01 -1.36813477e-01 3.08684379e-01 -5.07665455e-01
7.19156086e-01 4.91172910e-01 8.00178468e-01 -6.69179380e-01
9.38015342e-01 7.68489838e-02 -1.70671821e-01 -5.43257117e-01
-7.79163837e-02 6.47794679e-02 -4.97025788e-01 7.41579831e-01
-2.80386686e-01 6.30145788e-01 -5.69506705e-01 -2.31275260e-02
1.52815342e+00 -2.97715634e-01 6.37107909e-01 1.54705897e-01
7.01645315e-01 -2.08830222e-01 5.87357223e-01 6.03992760e-01
-7.46848285e-01 -5.17649472e-01 -3.70239854e-01 -1.07163393e+00
-8.15478802e-01 -1.01609945e+00 1.52398676e-01 1.08351660e+00
-1.21435247e-01 1.23643149e-02 -4.48062003e-01 -1.38485873e+00
3.65656227e-01 6.93759978e-01 -3.83279413e-01 -8.37035954e-01
-5.19280314e-01 -3.08800876e-01 9.00109410e-01 5.04746437e-01
7.66711175e-01 -9.71098661e-01 -1.02264094e+00 -1.00877710e-01
6.37190521e-01 -1.11739182e+00 -3.33887599e-02 5.13258040e-01
-7.08517194e-01 -1.42347383e+00 4.01713818e-01 -2.93799490e-01
5.37122071e-01 2.34907359e-01 5.57003558e-01 -1.73553172e-02
-3.73902768e-01 5.06584644e-01 -3.41348082e-01 -7.87290871e-01
-7.61918962e-01 8.71097371e-02 5.31247556e-01 2.47687697e-01
9.50401366e-01 -6.83279574e-01 -5.67296147e-01 3.44894499e-01
-1.22986770e+00 -8.77158940e-01 5.74987769e-01 4.38571543e-01
7.68567771e-02 5.55379212e-01 6.95489049e-01 -8.60951543e-01
4.40438181e-01 -7.35102594e-01 -8.80563796e-01 1.83078587e-01
-1.06673634e+00 -4.29572649e-02 1.00296509e+00 -8.00191998e-01
-9.93175209e-01 1.44115195e-01 -3.42236869e-02 -5.46796560e-01
-6.44180536e-01 -2.08802596e-01 -7.45817244e-01 -6.60232961e-01
5.17961442e-01 -3.19971710e-01 -2.22219490e-02 -4.09593642e-01
4.69074160e-01 1.00883472e+00 7.16425702e-02 -1.67973578e-01
9.44641888e-01 7.19479978e-01 1.42614678e-01 -3.40787381e-01
-8.15051615e-01 -3.48384500e-01 -2.87111789e-01 5.82298860e-02
8.17483008e-01 -6.26279294e-01 -1.49890769e+00 3.95077258e-01
-7.68017173e-01 4.68209505e-01 -3.20587039e-01 4.08378065e-01
1.01267092e-01 1.77066669e-01 -7.27792740e-01 -1.15956485e+00
-8.74133706e-01 -6.32915676e-01 4.14876193e-01 4.54691172e-01
-2.80583143e-01 -9.81882393e-01 1.32282719e-01 5.63515365e-01
6.54257655e-01 5.59711099e-01 7.67417490e-01 -1.69426942e+00
-6.58972383e-01 -1.15618849e+00 5.35506718e-02 4.46112007e-01
5.60417771e-01 -3.63451272e-01 -1.09101248e+00 -5.95887005e-01
5.28407931e-01 -2.73216277e-01 -1.21152848e-01 -9.94626805e-02
8.67039502e-01 -1.19838393e+00 -4.96239841e-01 5.38248837e-01
1.43112803e+00 6.35185122e-01 4.34250981e-01 5.44465661e-01
6.22287333e-01 3.66889924e-01 1.80921763e-01 5.88209331e-01
1.74859390e-01 -9.58800465e-02 9.98875380e-01 2.89371580e-01
7.82554030e-01 -4.54599053e-01 -1.48409475e-02 2.04122603e-01
6.25321746e-01 -1.47005498e-01 -6.12535715e-01 3.82294022e-02
-1.83491170e+00 -7.75752842e-01 2.71177202e-01 2.15654397e+00
-8.59086588e-02 3.52209151e-01 2.24599361e-01 3.28796923e-01
7.27362931e-01 1.00847818e-01 -1.10277569e+00 -8.99405122e-01
-1.82004133e-03 4.93019149e-02 1.05313563e+00 -2.55091965e-01
-1.36281264e+00 5.95619619e-01 6.38031912e+00 7.49186501e-02
-1.16786325e+00 3.15722466e-01 3.60484779e-01 1.38271272e-01
2.40540549e-01 1.73189908e-01 -3.76933366e-01 4.36344355e-01
1.37594712e+00 -3.99521440e-02 4.20191586e-01 1.57273483e+00
-2.51387786e-02 4.29172426e-01 -7.64519572e-01 8.71015608e-01
-3.34725231e-01 -1.11888170e+00 -1.82288066e-01 6.68998778e-01
4.79268759e-01 2.16140732e-01 1.36970729e-01 1.59863651e-01
8.18486869e-01 -7.58292735e-01 -6.91020191e-02 1.48391753e-01
1.31505921e-01 -9.40035224e-01 7.53730416e-01 3.15758556e-01
-9.32870507e-01 -8.46849501e-01 -1.03437670e-01 -3.50261271e-01
-1.30130962e-01 2.54520357e-01 -9.26240325e-01 4.36595827e-01
6.67987168e-01 1.90608948e-01 -3.47491980e-01 7.11580098e-01
-3.96539643e-02 5.61976492e-01 -5.64203620e-01 -2.38034591e-01
-1.30152926e-02 5.27040288e-02 4.46412951e-01 5.90599537e-01
-4.09649462e-02 6.84191734e-02 5.03860593e-01 6.38490498e-01
-3.10663909e-01 -1.08938828e-01 -1.12235475e+00 -4.40555543e-01
6.44672215e-01 1.42758167e+00 -6.10748172e-01 2.81829573e-02
-5.23614168e-01 9.01633620e-01 -6.96900114e-02 -9.98151749e-02
-5.80041647e-01 -1.67391747e-01 9.87773716e-01 1.37688473e-01
8.27051625e-02 2.88976375e-02 -5.96754253e-01 -1.16810942e+00
-2.74271339e-01 -8.51936579e-01 1.00683475e+00 -1.06293544e-01
-1.67416036e+00 5.92754841e-01 -3.95451874e-01 -1.20847142e+00
-1.04025170e-01 -5.46518505e-01 -7.46982753e-01 4.57264110e-02
-1.14847469e+00 -1.48625255e+00 3.79974246e-01 1.06522942e+00
-2.23724172e-01 -3.83132637e-01 1.19399476e+00 6.66941106e-02
-6.35967910e-01 5.89519322e-01 -7.24801496e-02 3.84475440e-01
2.43866779e-02 -6.97344005e-01 2.14287356e-01 9.92485762e-01
-4.98503074e-02 7.94322789e-01 3.69262755e-01 -9.32088673e-01
-1.97676015e+00 -1.19778097e+00 3.68996143e-01 -4.22348559e-01
7.57927716e-01 -3.75268400e-01 -8.22663486e-01 1.03512597e+00
1.72974646e-01 4.54533756e-01 1.12489653e+00 -7.30030313e-02
-8.82016659e-01 -4.79923934e-01 -2.19264841e+00 3.64547938e-01
3.47120792e-01 -5.99633574e-01 -4.28707600e-01 1.89401418e-01
1.07985044e+00 2.51109689e-01 -1.03674114e+00 2.87169039e-01
5.44561148e-01 -7.46328354e-01 8.38084280e-01 -1.06665421e+00
-7.32748985e-01 -1.72224388e-01 -4.11646873e-01 -5.47533929e-01
-9.82288495e-02 -9.73541081e-01 -6.43155277e-01 1.26646304e+00
1.54490337e-01 -1.27815330e+00 1.14604878e+00 1.39215243e+00
4.66547638e-01 -1.37957752e-01 -1.08515012e+00 -7.96415329e-01
-1.44716859e-01 -5.03643334e-01 1.21096790e+00 1.05312908e+00
1.62547573e-01 4.32992019e-02 -4.62198406e-01 5.93702614e-01
9.88243937e-01 -2.23493993e-01 4.88668263e-01 -1.40467417e+00
7.45581239e-02 1.54360160e-01 -7.74231493e-01 1.30168989e-01
2.06042469e-01 -4.72443163e-01 -7.08489120e-01 -6.19003177e-01
-4.52487648e-01 -4.70556527e-01 -1.10237825e+00 9.16953564e-01
5.91538310e-01 1.96027502e-01 1.92896605e-01 4.87882383e-02
-7.86710262e-01 1.82894528e-01 2.86688358e-01 -3.72901917e-01
-2.86268294e-01 4.23674375e-01 -9.76357341e-01 8.48155022e-01
1.27494299e+00 -6.52845621e-01 -2.51900494e-01 5.62615804e-02
1.79025352e-01 -5.81791159e-03 3.69132429e-01 -1.15825748e+00
4.29511428e-01 -2.26733699e-01 4.31785017e-01 -3.27728063e-01
1.42721459e-01 -1.81405723e+00 3.84486854e-01 8.55604351e-01
-2.89409012e-02 8.73316750e-02 1.71777830e-01 6.88517094e-01
2.56765187e-01 2.88031191e-01 8.32003653e-01 -8.54508951e-02
-1.85102120e-01 5.01254141e-01 -2.93467820e-01 -6.34617805e-01
1.61308515e+00 -2.76617527e-01 -4.85108405e-01 -5.83976321e-02
-5.70855319e-01 6.60705417e-02 7.61401117e-01 6.48088276e-01
5.83544910e-01 -1.10756814e+00 2.57867187e-01 5.75592041e-01
2.33560711e-01 -6.10728264e-01 -2.72112936e-02 1.46727219e-01
-1.55978305e-02 4.05060053e-01 -5.17067134e-01 -1.02201670e-01
-1.27494812e+00 1.58253467e+00 1.73784241e-01 -3.02343220e-01
-6.44014776e-01 1.43127799e-01 -8.04901779e-01 -4.65426624e-01
4.36633766e-01 -9.68860928e-03 2.30459258e-01 -4.89554971e-01
8.31286371e-01 6.17670715e-01 1.62815765e-01 -5.01175344e-01
-9.11967754e-01 -1.63578823e-01 -3.13390613e-01 3.40758681e-01
1.13229060e+00 -1.78477317e-01 -6.68455884e-02 -8.71651992e-02
1.43636453e+00 -7.60494471e-02 -9.71408725e-01 -2.87166312e-02
3.59859765e-01 -6.35409296e-01 3.90270688e-02 -1.31692958e+00
-1.19196010e+00 2.67531365e-01 9.91882384e-01 7.65167773e-01
1.26571023e+00 -2.62278527e-01 1.24946082e+00 5.88842392e-01
7.95416534e-01 -7.37301826e-01 -1.71672687e-01 -1.03521481e-01
-3.32985699e-01 -1.47519636e+00 -1.41034007e-01 -2.29225621e-01
-2.73894012e-01 8.81841481e-01 4.09699738e-01 -2.86785483e-01
1.05332160e+00 7.12849200e-01 7.81548470e-02 -2.95135491e-02
-4.19645280e-01 3.26630443e-01 -2.85484105e-01 1.00909913e+00
-5.00478208e-01 3.17984700e-01 9.41251740e-02 6.20106995e-01
2.89596736e-01 1.53183684e-01 1.53249308e-01 1.51161742e+00
-3.27496976e-01 -1.60458469e+00 -5.40992916e-01 5.17859399e-01
-9.14891779e-01 4.32163000e-01 -5.33656180e-01 3.15480739e-01
2.83938169e-01 1.37799442e+00 -5.20076275e-01 -9.11289155e-01
7.53259636e-04 1.65357903e-01 -1.31525770e-01 -1.15609013e-01
-1.02518630e+00 -4.71522152e-01 -1.78594306e-01 -7.80048609e-01
-2.35309914e-01 -2.49302015e-01 -1.23486924e+00 -4.61175948e-01
-3.45810115e-01 1.86112523e-02 1.09192705e+00 8.77862930e-01
8.09470117e-01 1.79875597e-01 1.07280660e+00 -1.76510483e-01
-8.23094547e-01 -3.74963313e-01 -4.26211149e-01 3.62217247e-01
3.89157295e-01 -2.87730008e-01 -5.35552919e-01 -5.54281712e-01] | [5.366448879241943, 7.141229629516602] |
b3ecf7d5-08b0-40bb-9254-e26ff1426c95 | semi-supervised-deep-quick-instance-detection | 2101.06405 | null | https://arxiv.org/abs/2101.06405v1 | https://arxiv.org/pdf/2101.06405v1.pdf | Semi Supervised Deep Quick Instance Detection and Segmentation | In this paper, we present a semi supervised deep quick learning framework for instance detection and pixel-wise semantic segmentation of images in a dense clutter of items. The framework can quickly and incrementally learn novel items in an online manner by real-time data acquisition and generating corresponding ground truths on its own. To learn various combinations of items, it can synthesize cluttered scenes, in real time. The overall approach is based on the tutor-child analogy in which a deep network (tutor) is pretrained for class-agnostic object detection which generates labeled data for another deep network (child). The child utilizes a customized convolutional neural network head for the purpose of quick learning. There are broadly four key components of the proposed framework semi supervised labeling, occlusion aware clutter synthesis, a customized convolutional neural network head, and instance detection. The initial version of this framework was implemented during our participation in Amazon Robotics Challenge (ARC), 2017. Our system was ranked 3rd, 4th and 5th worldwide in pick, stow-pick and stow task respectively. The proposed framework is an improved version over ARC17 where novel features such as instance detection and online learning has been added. | ['L. Behera', 'Ashish Kumar'] | 2021-01-16 | null | null | null | null | ['class-agnostic-object-detection'] | ['computer-vision'] | [ 3.53252739e-01 3.67736876e-01 3.01590413e-01 -6.71593904e-01
-6.96852684e-01 -8.59038770e-01 5.64583123e-01 2.48566344e-01
-6.49841189e-01 8.44268739e-01 -4.63023424e-01 1.50301661e-02
-6.87269121e-02 -7.61725008e-01 -1.22312069e+00 -2.95262963e-01
-2.22391620e-01 8.99244905e-01 6.65775776e-01 1.02958135e-01
2.05014184e-01 7.57721841e-01 -1.87882483e+00 5.45927703e-01
7.48133957e-01 1.25765657e+00 6.53737307e-01 8.92784715e-01
-2.76572138e-01 6.91943288e-01 -5.91841638e-01 -2.83025414e-01
8.42092335e-01 1.89587772e-01 -9.95145023e-01 6.25209510e-01
7.20592678e-01 -4.97365475e-01 1.57735139e-01 9.01921570e-01
5.08750260e-01 6.65730387e-02 6.45929813e-01 -1.41991055e+00
-3.22717249e-01 8.80380571e-01 -5.43873549e-01 9.02753249e-02
1.33029163e-01 4.17247176e-01 6.17992401e-01 -1.27885175e+00
7.42109597e-01 1.06383407e+00 6.65884495e-01 3.14602137e-01
-9.02745068e-01 -7.14184999e-01 4.02848721e-01 1.66653395e-01
-1.08008552e+00 -1.20910674e-01 6.96601987e-01 -4.52901542e-01
7.75596321e-01 -1.14427455e-01 7.21682131e-01 9.49392498e-01
-1.62162200e-01 1.36477816e+00 1.33446288e+00 -3.02898705e-01
5.34520149e-01 4.79946882e-01 2.31556743e-01 7.86886036e-01
4.27261025e-01 1.51796803e-01 -4.00611520e-01 3.51601154e-01
7.01312959e-01 4.77999961e-03 3.89915377e-01 -6.23861611e-01
-1.15232825e+00 4.84483093e-01 7.78356075e-01 -1.58625007e-01
-6.59116745e-01 2.30930924e-01 4.10038382e-01 7.11873397e-02
1.73937261e-01 4.01146591e-01 -6.16544068e-01 3.75529617e-01
-1.06980956e+00 6.82365537e-01 7.03044176e-01 1.57696605e+00
8.37249994e-01 8.14649537e-02 -3.07354152e-01 5.27517259e-01
1.83803722e-01 2.63691753e-01 4.68642175e-01 -9.45745885e-01
3.18934202e-01 4.64407176e-01 5.11512876e-01 -3.66076112e-01
-5.95500767e-01 -7.99458385e-01 -2.54968196e-01 5.34597218e-01
3.35627288e-01 -4.38376635e-01 -1.43657565e+00 1.18665493e+00
7.59536088e-01 2.01141968e-01 -1.49121717e-01 8.67496014e-01
1.00092745e+00 3.10707122e-01 1.77984476e-01 2.44375423e-01
1.32474411e+00 -1.52991462e+00 -3.87737185e-01 -3.56436253e-01
1.08597994e-01 -8.35315406e-01 8.73246968e-01 7.45641112e-01
-1.18190885e+00 -9.50665355e-01 -1.34248126e+00 5.94419390e-02
-9.43946004e-01 7.56449476e-02 7.62569070e-01 4.79676843e-01
-8.99208248e-01 5.43928862e-01 -4.85680819e-01 -3.78090411e-01
8.41904819e-01 3.85860413e-01 -1.09864414e-01 -2.77182221e-01
-6.91430449e-01 7.30140269e-01 8.95852447e-01 1.32387355e-01
-1.46703041e+00 -5.89168727e-01 -6.61429644e-01 -3.80207270e-01
7.14560628e-01 -5.56654572e-01 1.50114787e+00 -1.39656281e+00
-1.41271865e+00 1.02623999e+00 4.42947865e-01 -1.01138699e+00
9.79413807e-01 -5.13725579e-01 -1.77481920e-01 1.83282048e-01
4.04462099e-01 1.21237504e+00 1.09257102e+00 -1.52527809e+00
-1.19034743e+00 -2.72718221e-01 2.31575549e-01 1.60721213e-01
5.24783075e-01 -2.05519110e-01 -3.40052098e-01 -6.66953385e-01
-5.02960905e-02 -8.55987787e-01 -3.40328276e-01 1.81398913e-01
-6.39909565e-01 -2.32991979e-01 8.84676516e-01 -2.54964322e-01
3.78175139e-01 -1.75106299e+00 -2.46494308e-01 1.98335368e-02
1.25313610e-01 3.38133126e-01 -1.95533887e-01 1.09030865e-01
-7.35304132e-02 -4.58311439e-01 -2.81609654e-01 -3.93360764e-01
5.97554371e-02 2.20517844e-01 -1.06245898e-01 3.30071390e-01
6.66445255e-01 1.01462805e+00 -1.07225955e+00 -4.81462836e-01
5.03569245e-01 9.67607051e-02 -2.95080721e-01 3.03288043e-01
-7.79686451e-01 1.21215977e-01 -1.76358817e-03 9.95059192e-01
1.08846366e+00 8.37338120e-02 -2.69271463e-01 -2.10700363e-01
-3.56346190e-01 -1.39784291e-01 -1.76433384e+00 1.85639501e+00
-3.87784243e-01 4.77962613e-01 7.48564973e-02 -1.02839863e+00
8.44491243e-01 -5.12895696e-02 9.52061638e-02 -5.06825805e-01
1.95879713e-01 3.38051349e-01 -3.13376725e-01 -6.24703526e-01
5.92008770e-01 1.95428550e-01 7.30945310e-03 2.99707949e-01
5.78175843e-01 -1.08514406e-01 5.14460623e-01 2.16375589e-01
8.94416094e-01 7.66622961e-01 1.75198808e-01 -3.89889270e-01
2.85198003e-01 5.05741835e-01 1.79016635e-01 1.01444018e+00
-4.28899765e-01 6.59048319e-01 5.06037883e-02 -8.39083731e-01
-1.09753799e+00 -1.27710485e+00 -7.33104795e-02 1.26231253e+00
2.73338616e-01 2.61882871e-01 -8.55001152e-01 -1.00684810e+00
4.48765699e-03 5.30687332e-01 -7.46674836e-01 3.13922554e-01
-5.78053355e-01 -2.56464779e-01 2.66046077e-01 6.84820414e-01
8.98333430e-01 -1.62153482e+00 -1.33342361e+00 5.02394140e-01
2.69744873e-01 -1.17054474e+00 -9.44479331e-02 8.03857565e-01
-6.29524112e-01 -1.27977777e+00 -9.09820437e-01 -1.04958975e+00
8.11064482e-01 3.00354809e-01 1.14833045e+00 -4.47834730e-01
-9.27538753e-01 2.82813072e-01 -3.81392121e-01 -1.09948659e+00
5.53325303e-02 2.65293047e-02 4.59785610e-02 6.74617067e-02
8.76712278e-02 -2.80198574e-01 -8.45497668e-01 1.94129780e-01
-1.00501084e+00 6.47784844e-02 9.62120831e-01 6.13363206e-01
8.81464958e-01 -5.85182086e-02 6.03250086e-01 -1.45885050e+00
2.73890525e-01 -5.04978597e-01 -8.80721271e-01 1.25524402e-01
-3.26660454e-01 1.48293719e-01 5.81818640e-01 -4.88970160e-01
-1.15845978e+00 7.93975711e-01 -1.58907890e-01 -3.09427261e-01
-5.97640693e-01 -8.33558068e-02 -6.46510720e-02 -4.73744124e-02
7.44770825e-01 2.75088828e-02 -7.26331174e-01 -5.28556764e-01
7.34724045e-01 5.89140892e-01 8.13579619e-01 -6.37589037e-01
7.40456700e-01 4.52597409e-01 -1.52703553e-01 -3.70857984e-01
-1.13993633e+00 -5.70809066e-01 -1.09717238e+00 -4.27128434e-01
6.88448727e-01 -1.00024116e+00 -3.36738944e-01 6.59565508e-01
-1.08872426e+00 -6.12480640e-01 -7.44537950e-01 -6.13099970e-02
-5.12849569e-01 -1.20559663e-01 -1.15518342e-03 -9.35943842e-01
-4.00474548e-01 -9.58823025e-01 1.21133566e+00 4.23418432e-01
1.80961534e-01 -5.14989853e-01 -2.98429668e-01 1.74044564e-01
3.93968910e-01 6.88895643e-01 3.16456258e-01 -8.41786504e-01
-8.13110590e-01 -3.02123904e-01 -6.01984739e-01 4.91373211e-01
-1.97304502e-01 -1.71489894e-01 -1.17364621e+00 5.14800660e-03
-3.94466966e-01 -5.53660035e-01 8.25502813e-01 1.70261458e-01
1.22343278e+00 -2.83010304e-01 -2.20549360e-01 5.52277803e-01
1.82291842e+00 1.97819516e-01 2.25045010e-01 4.91630942e-01
4.91344541e-01 6.68824673e-01 9.66446221e-01 4.84341174e-01
3.33745331e-01 3.92752528e-01 8.76117766e-01 -1.73107997e-01
-4.85067934e-01 -2.18676805e-01 7.96308964e-02 -9.12299678e-02
2.17572674e-01 1.03685163e-01 -8.61229062e-01 8.53898883e-01
-1.77856493e+00 -5.56393862e-01 -1.92482442e-01 1.98742473e+00
5.05901754e-01 5.46401381e-01 5.79789519e-01 1.65603638e-01
5.33951461e-01 -3.90979558e-01 -8.70548546e-01 -4.38075930e-01
9.11209360e-03 6.76122069e-01 7.71736681e-01 3.39775026e-01
-1.29549754e+00 1.08270228e+00 5.52180195e+00 4.59929377e-01
-9.93188679e-01 2.67807752e-01 4.97429729e-01 -3.39466962e-03
2.59519070e-01 -2.60964245e-01 -9.00425911e-01 2.64000267e-01
5.70883930e-01 1.32460177e-01 3.29664141e-01 1.32286799e+00
-5.00998041e-03 -4.32478189e-01 -1.13837636e+00 7.70582616e-01
2.90962849e-02 -1.34892559e+00 1.27479702e-01 -5.23836911e-01
8.69552433e-01 1.95784792e-01 1.70428693e-01 4.94824022e-01
5.24614811e-01 -7.20773458e-01 1.18083870e+00 4.57032144e-01
6.10369921e-01 -8.74361098e-01 7.01412559e-01 3.78599852e-01
-1.16096687e+00 -3.60352725e-01 -2.85033673e-01 1.58739448e-01
-2.60913372e-01 4.63710517e-01 -1.18209112e+00 3.50299746e-01
8.94838452e-01 2.55481273e-01 -8.02116275e-01 1.42887211e+00
-2.35583603e-01 2.23122343e-01 -3.89219940e-01 -7.30827823e-02
6.63880825e-01 3.59709412e-01 4.28060085e-01 1.68918490e+00
-1.96752697e-01 -1.10041134e-01 4.17000473e-01 9.63983536e-01
-1.03043243e-01 -9.10123512e-02 -3.67397547e-01 3.84485781e-01
3.95982802e-01 1.59350479e+00 -1.20990682e+00 -7.58392572e-01
-6.06316291e-02 1.23143744e+00 3.67387354e-01 1.08383678e-01
-8.07746589e-01 -7.12793946e-01 -1.83515740e-03 3.27175438e-01
5.20484865e-01 4.14760299e-02 -4.19178605e-01 -6.67269528e-01
-4.31696139e-02 -4.89502072e-01 4.01482105e-01 -1.00410473e+00
-1.19110572e+00 7.11731315e-01 5.08492775e-02 -1.06271183e+00
-8.66467804e-02 -8.49934101e-01 -5.02053261e-01 5.14105082e-01
-1.63853264e+00 -1.33567655e+00 -7.53901064e-01 7.91972160e-01
1.36618614e+00 -6.53447909e-03 5.14392972e-01 1.77831396e-01
-3.77063900e-01 4.34554845e-01 -2.48220846e-01 -5.48949391e-02
4.28179532e-01 -1.70529962e+00 3.64131391e-01 9.19591904e-01
1.46132290e-01 2.11539745e-01 7.62700856e-01 -5.56514859e-01
-1.14054370e+00 -1.78398407e+00 2.17036903e-01 -2.47043520e-01
3.87260765e-01 -6.58473074e-01 -4.43367809e-01 6.02726161e-01
2.07687631e-01 2.44656771e-01 1.28911674e-01 -5.67341805e-01
-2.16189012e-01 -2.59429991e-01 -1.64550853e+00 2.35628307e-01
9.86679256e-01 1.04537435e-01 -5.09595335e-01 7.97330439e-01
8.06448817e-01 -6.30455375e-01 -3.22283983e-01 3.27736646e-01
5.35884321e-01 -9.60515320e-01 8.28623295e-01 -4.76273060e-01
4.25674260e-01 -5.38589776e-01 -9.04797539e-02 -8.14866602e-01
-1.38114288e-01 -6.56584084e-01 -2.70370841e-01 9.49617088e-01
5.06421566e-01 -2.03093454e-01 1.06216800e+00 4.49023485e-01
-3.80626887e-01 -7.60656655e-01 -4.62130398e-01 -8.40611041e-01
-3.65709305e-01 -5.78763425e-01 5.19027531e-01 4.63615268e-01
-6.46092117e-01 1.30475566e-01 -7.13470951e-02 3.43910187e-01
1.07317960e+00 1.71168908e-01 9.62684035e-01 -1.07264149e+00
-3.74358177e-01 -1.14856400e-01 -2.99961358e-01 -7.47792542e-01
-4.29598033e-01 -7.21627414e-01 3.97016019e-01 -1.72358584e+00
4.70279250e-03 -5.50170064e-01 -2.71376699e-01 4.18129027e-01
-2.17212252e-02 3.22384655e-01 3.51203710e-01 -4.61804122e-02
-9.90764976e-01 -1.24686368e-01 8.89674366e-01 -1.77534282e-01
-1.95641652e-01 3.17966104e-01 -4.82800335e-01 7.47078717e-01
7.32436299e-01 -4.41713780e-01 -2.39907339e-01 -4.10520971e-01
-6.77477941e-02 -4.48284984e-01 5.60782611e-01 -1.29724658e+00
2.18634799e-01 1.32267877e-01 8.22439015e-01 -1.11300290e+00
1.00646071e-01 -1.03336287e+00 4.95224725e-03 5.88641107e-01
-3.05415809e-01 -9.29957107e-02 3.71792406e-01 6.19434536e-01
1.22165114e-01 -4.62591231e-01 9.50063288e-01 -7.62320697e-01
-1.17429745e+00 3.04934412e-01 -1.62222534e-01 -1.49426788e-01
1.40574396e+00 -4.24452096e-01 2.93223243e-02 1.77535430e-01
-9.90249336e-01 3.29331398e-01 2.39895172e-02 4.58435953e-01
7.12987661e-01 -1.00723195e+00 -5.50409615e-01 1.31851450e-01
3.03525448e-01 6.77342772e-01 1.30558327e-01 2.57031560e-01
-8.15460205e-01 2.06792414e-01 -5.12406349e-01 -6.81658983e-01
-8.02387297e-01 7.89277315e-01 2.94469088e-01 -2.31165692e-01
-5.35094023e-01 1.30008233e+00 6.74299076e-02 -4.28877264e-01
7.09401846e-01 -3.35995078e-01 -5.83462138e-03 -5.22000864e-02
5.35952628e-01 4.50858206e-01 1.39411956e-01 -1.44228533e-01
-2.28771061e-01 2.47661456e-01 -2.46271715e-01 -9.49903801e-02
1.49140906e+00 -1.95158273e-03 2.65100658e-01 4.55256373e-01
8.23466897e-01 -5.11114299e-01 -1.59340322e+00 -3.18291754e-01
2.24678397e-01 -1.68859959e-01 -1.87311113e-01 -1.31373119e+00
-9.28612769e-01 7.48479843e-01 1.01974225e+00 -6.69550225e-02
1.07671738e+00 -1.03330128e-01 7.45440364e-01 5.20029247e-01
6.65840685e-01 -1.36971712e+00 3.90735388e-01 2.93255359e-01
9.02224004e-01 -1.21298993e+00 -8.14307034e-02 -2.73065567e-01
-6.90695763e-01 1.12865770e+00 1.02347279e+00 -7.05288231e-01
5.68130910e-01 5.34961343e-01 -1.46273091e-01 -3.01392466e-01
-4.89570916e-01 -4.44022834e-01 9.29415897e-02 1.02188396e+00
-1.10647365e-01 2.04248235e-01 2.13159934e-01 6.57209992e-01
-1.16420045e-01 1.46042675e-01 5.78782737e-01 1.13762891e+00
-7.60022879e-01 -6.10319912e-01 -3.19191635e-01 4.95688051e-01
-2.44436219e-01 7.07329288e-02 -4.01682079e-01 8.20976853e-01
8.09231102e-01 5.55266559e-01 -4.12659347e-02 -3.85038145e-02
5.69196284e-01 -1.13590211e-02 5.17080069e-01 -9.38132823e-01
-7.27799654e-01 -2.14658722e-01 1.18817994e-02 -7.23674357e-01
-4.00052756e-01 -5.27554572e-01 -1.26625967e+00 3.11545968e-01
-3.05384338e-01 -3.13525081e-01 1.23697758e+00 9.26052988e-01
1.47581801e-01 6.42158628e-01 4.26872611e-01 -1.33294451e+00
-1.62922233e-01 -8.79353762e-01 -5.10687113e-01 1.79928437e-01
3.41630578e-01 -6.98130965e-01 2.25742966e-01 3.24016869e-01] | [9.401999473571777, 0.30036288499832153] |
1c526b4d-31d4-4b06-855e-77d33bd2bd3f | re-thinking-co-salient-object-detection | 2007.03380 | null | https://arxiv.org/abs/2007.03380v4 | https://arxiv.org/pdf/2007.03380v4.pdf | Re-thinking Co-Salient Object Detection | In this paper, we conduct a comprehensive study on the co-salient object detection (CoSOD) problem for images. CoSOD is an emerging and rapidly growing extension of salient object detection (SOD), which aims to detect the co-occurring salient objects in a group of images. However, existing CoSOD datasets often have a serious data bias, assuming that each group of images contains salient objects of similar visual appearances. This bias can lead to the ideal settings and effectiveness of models trained on existing datasets, being impaired in real-life situations, where similarities are usually semantic or conceptual. To tackle this issue, we first introduce a new benchmark, called CoSOD3k in the wild, which requires a large amount of semantic context, making it more challenging than existing CoSOD datasets. Our CoSOD3k consists of 3,316 high-quality, elaborately selected images divided into 160 groups with hierarchical annotations. The images span a wide range of categories, shapes, object sizes, and backgrounds. Second, we integrate the existing SOD techniques to build a unified, trainable CoSOD framework, which is long overdue in this field. Specifically, we propose a novel CoEG-Net that augments our prior model EGNet with a co-attention projection strategy to enable fast common information learning. CoEG-Net fully leverages previous large-scale SOD datasets and significantly improves the model scalability and stability. Third, we comprehensively summarize 40 cutting-edge algorithms, benchmarking 18 of them over three challenging CoSOD datasets (iCoSeg, CoSal2015, and our CoSOD3k), and reporting more detailed (i.e., group-level) performance analysis. Finally, we discuss the challenges and future works of CoSOD. We hope that our study will give a strong boost to growth in the CoSOD community. The benchmark toolbox and results are available on our project page at http://dpfan.net/CoSOD3K/. | ['Ming-Ming Cheng', 'Ge-Peng Ji', 'Tengpeng Li', 'Deng-Ping Fan', 'Huazhu Fu', 'Dingwen Zhang', 'Zheng Lin', 'Jianbing Shen'] | 2020-07-07 | null | null | null | null | ['co-saliency-detection'] | ['computer-vision'] | [ 5.67008890e-02 -1.43742412e-01 -2.92813540e-01 -6.09444752e-02
-5.41360497e-01 -2.66119868e-01 6.30034626e-01 1.72244478e-02
-2.39971235e-01 4.07981843e-01 6.11881435e-01 1.13446638e-01
1.22604772e-01 -4.37547773e-01 -7.19833374e-01 -6.10816658e-01
1.63656637e-01 -4.15656716e-02 5.89273751e-01 -1.78044409e-01
2.60150462e-01 4.10120189e-01 -1.88092864e+00 1.70411333e-01
7.17605591e-01 1.09319305e+00 7.11588740e-01 2.99686879e-01
3.06444615e-01 6.62383318e-01 -3.08714211e-01 -2.88093686e-01
2.50108212e-01 -7.23073259e-02 -8.30707550e-01 6.61331713e-02
9.78680551e-01 -2.14244410e-01 -4.39826429e-01 1.28072298e+00
6.60617590e-01 1.07504100e-01 4.49299246e-01 -1.52585936e+00
-9.50957716e-01 6.12360477e-01 -6.38453484e-01 4.44822669e-01
-5.90511523e-02 1.85331047e-01 1.38591492e+00 -1.37309909e+00
6.98867559e-01 1.25784600e+00 6.90007389e-01 3.71610373e-01
-1.12891233e+00 -6.47463083e-01 5.85303485e-01 5.82847059e-01
-1.25967181e+00 -2.77237982e-01 8.93952429e-01 -3.54738832e-01
8.70307922e-01 3.00509989e-01 7.78050303e-01 1.25780201e+00
-6.98922947e-02 1.36727762e+00 8.17226887e-01 -2.69553602e-01
1.19301632e-01 -7.61732459e-02 2.06197470e-01 3.84101868e-01
4.15804863e-01 -4.29181829e-02 -5.87161183e-01 6.20384924e-02
6.04756594e-01 4.12850648e-01 -3.63541335e-01 -7.47792363e-01
-1.44973969e+00 7.04320669e-01 9.21779692e-01 1.95633501e-01
-3.78062755e-01 -4.31086868e-02 3.53535771e-01 -1.71604887e-01
4.67531800e-01 5.20436287e-01 -3.13613236e-01 1.80249676e-01
-7.15491951e-01 5.05596399e-01 3.04837912e-01 1.08169520e+00
5.92220008e-01 -1.69826657e-01 -5.04249871e-01 1.00438154e+00
1.62979737e-01 4.80229676e-01 5.06425500e-01 -7.82114565e-01
2.53741533e-01 7.40734100e-01 1.43597513e-01 -1.34773409e+00
-5.12987614e-01 -5.93767941e-01 -6.95855379e-01 9.97704118e-02
7.68347234e-02 1.67596817e-01 -9.40413415e-01 1.72468078e+00
3.71929705e-01 1.99691102e-01 -2.23835483e-01 1.31736684e+00
1.15649045e+00 5.63575804e-01 2.29806647e-01 1.55348584e-01
1.40690947e+00 -1.59090388e+00 -4.91567463e-01 -5.03384829e-01
3.17134559e-01 -8.77397478e-01 1.28960598e+00 2.06088603e-01
-9.81556177e-01 -6.46435976e-01 -1.01478350e+00 -3.88018787e-01
-4.71984088e-01 2.41520792e-01 8.92774105e-01 -1.07129477e-01
-1.00792921e+00 2.94870406e-01 -6.24538958e-01 -5.64527929e-01
6.65078878e-01 1.43144298e-02 -5.46852909e-02 -1.88609779e-01
-9.38830554e-01 7.05487847e-01 4.20965791e-01 6.31105229e-02
-1.10596049e+00 -8.86832952e-01 -8.97873759e-01 8.65827203e-02
6.11271977e-01 -5.49522221e-01 1.26692057e+00 -9.48172152e-01
-7.02772200e-01 1.12171316e+00 -2.04940364e-02 -3.65683466e-01
4.18146044e-01 -5.91077685e-01 -5.37288308e-01 1.44557953e-01
4.72865671e-01 1.02178824e+00 8.85935545e-01 -1.21812344e+00
-7.46604741e-01 -1.39174059e-01 1.28609523e-01 3.70091736e-01
-4.82256174e-01 2.91057587e-01 -8.68296683e-01 -9.47858572e-01
-1.37044145e-02 -8.47429872e-01 -1.74021900e-01 2.38203123e-01
-6.44980073e-01 -3.98530036e-01 9.20878887e-01 -3.58468920e-01
1.13166428e+00 -2.52097249e+00 1.44476473e-01 -4.76992339e-01
5.76674640e-01 4.49875414e-01 -2.68714339e-01 2.86673188e-01
-2.15528131e-01 -1.45757824e-01 -1.25545889e-01 -6.46657586e-01
1.20018139e-01 -1.12912938e-01 -4.89069134e-01 2.64451295e-01
2.87792653e-01 1.12651861e+00 -1.03438115e+00 -4.83881414e-01
2.72507161e-01 4.49711680e-01 -4.22648251e-01 1.52519196e-01
-1.17874429e-01 -1.33968264e-01 -3.98891896e-01 8.13710451e-01
7.66384602e-01 -7.21509993e-01 -2.74648726e-01 -3.96769345e-01
-1.38835877e-01 2.21957624e-01 -1.19012284e+00 1.54893434e+00
-4.62177545e-02 7.14043140e-01 -6.48019239e-02 -9.32363749e-01
8.56148362e-01 -1.69269621e-01 4.21470553e-01 -7.60458469e-01
2.11886708e-02 2.28729874e-01 -2.86457092e-01 -2.85352826e-01
8.89310777e-01 3.23394358e-01 1.10692352e-01 1.44683450e-01
8.13146979e-02 1.19839080e-01 3.16215605e-01 5.04276872e-01
7.70469427e-01 -6.95777237e-02 3.80066693e-01 -5.29841721e-01
2.54951656e-01 3.94886509e-02 8.46466362e-01 7.09647417e-01
-5.32552361e-01 1.12126744e+00 3.71132195e-01 -5.42574942e-01
-9.08423543e-01 -1.05260885e+00 -2.62569457e-01 1.16033924e+00
7.71186352e-01 -4.50115979e-01 -5.04399657e-01 -8.67131472e-01
2.90609628e-01 3.80896658e-01 -8.20949018e-01 -1.90953672e-01
-3.37457389e-01 -8.91084313e-01 -1.95303205e-02 7.71611810e-01
5.90584159e-01 -1.48314297e+00 -4.64217186e-01 9.96237695e-02
-1.56473726e-01 -1.18900609e+00 -8.41091216e-01 9.07185972e-02
-4.84883428e-01 -1.09874105e+00 -8.68295312e-01 -1.08115566e+00
5.60399592e-01 1.05805397e+00 1.48229492e+00 1.39070764e-01
-4.28986192e-01 2.22539857e-01 -4.99664068e-01 -7.73421943e-01
3.01131964e-01 6.27988875e-02 1.69366911e-01 1.68145858e-02
6.68568552e-01 -2.77525544e-01 -9.39853370e-01 6.15588009e-01
-7.25152493e-01 4.15015519e-01 5.26322603e-01 7.72888780e-01
7.65428483e-01 -2.79340565e-01 6.38057828e-01 -5.10435462e-01
1.47478491e-01 -5.88312089e-01 -6.34882808e-01 1.01614624e-01
-5.67673266e-01 -3.45250219e-01 4.45821762e-01 -2.81333089e-01
-7.71362126e-01 -1.56589076e-01 1.17969684e-01 -6.32310212e-01
3.08211949e-02 2.32461035e-01 -2.51197726e-01 1.01661393e-02
5.43458223e-01 2.34162256e-01 -3.90418321e-01 -7.16839373e-01
1.43796444e-01 5.21744430e-01 7.50710666e-01 -3.61628413e-01
6.89838886e-01 5.12659431e-01 -4.09373105e-01 -5.90259671e-01
-1.39297032e+00 -7.07796454e-01 -3.85920376e-01 -1.56118779e-03
7.25200474e-01 -1.26088190e+00 -1.32184699e-01 7.13911235e-01
-9.32504654e-01 -4.01230574e-01 -3.38915646e-01 4.56709206e-01
-2.11038798e-01 2.83166230e-01 -4.53540653e-01 -3.27415287e-01
-4.31539536e-01 -1.16315496e+00 1.35917628e+00 5.26112616e-01
5.65025248e-02 -8.29090059e-01 -1.63585246e-01 2.69356459e-01
3.34195971e-01 2.10867450e-02 3.53723615e-01 -4.59361732e-01
-7.55298615e-01 2.35099524e-01 -7.30389237e-01 3.06579083e-01
7.62681067e-02 -1.76421907e-02 -9.63083208e-01 -4.49885339e-01
-4.44776982e-01 -5.10745287e-01 1.24150646e+00 5.23047984e-01
1.42337465e+00 3.37922797e-02 -5.03082871e-01 7.58601010e-01
1.31651461e+00 -1.35493264e-01 3.46827745e-01 5.45754611e-01
8.28162849e-01 4.32017744e-01 9.35105741e-01 3.12713742e-01
5.22342265e-01 9.82119441e-01 7.58338451e-01 -6.06064379e-01
-5.30136704e-01 -3.55487853e-01 2.60552645e-01 5.84212065e-01
-3.33876461e-02 -2.17430875e-01 -7.13370919e-01 1.00830805e+00
-1.97545683e+00 -8.87967885e-01 -9.62814242e-02 1.81919372e+00
7.29565024e-01 -8.50195903e-03 3.16520363e-01 -1.53028876e-01
9.91857469e-01 4.52122450e-01 -7.29968250e-01 1.47626966e-01
-4.82388526e-01 -2.06270516e-01 2.93329954e-01 7.66606107e-02
-1.50136292e+00 9.70327675e-01 6.06262398e+00 9.05548155e-01
-1.18732345e+00 -3.52571122e-02 6.18671656e-01 -2.57086724e-01
-1.82133004e-01 -2.25333303e-01 -1.15512037e+00 6.75000072e-01
1.07635044e-01 -2.44207084e-01 8.07387158e-02 1.25229502e+00
1.26902476e-01 -2.60545425e-02 -8.70288312e-01 1.22012281e+00
3.13674420e-01 -1.50774419e+00 -6.06394149e-02 -1.52445197e-01
9.79916990e-01 5.60905457e-01 9.68384519e-02 3.37626606e-01
3.42215776e-01 -8.75876188e-01 8.55649710e-01 7.99390674e-02
4.65306044e-01 -4.11664814e-01 6.43988252e-01 -4.35965359e-02
-1.27705097e+00 -1.05090998e-01 -6.89333975e-01 -1.63240246e-02
4.42842692e-02 7.28053451e-01 -4.28014994e-01 4.73902345e-01
1.37820649e+00 1.37440693e+00 -8.83071065e-01 1.45521593e+00
-2.34700128e-01 5.50061822e-01 -2.47585222e-01 -2.13039736e-03
3.53177577e-01 9.66439471e-02 5.93092561e-01 1.24069941e+00
1.49831414e-01 -2.14197651e-01 2.73109347e-01 9.31456983e-01
-3.08914691e-01 6.14058152e-02 -3.34768683e-01 1.23399235e-01
4.79642719e-01 1.59568083e+00 -7.46747196e-01 -4.28093225e-01
-6.66286767e-01 8.81313145e-01 3.68704081e-01 3.44920814e-01
-8.51513803e-01 -2.35506788e-01 8.86827230e-01 -3.26366839e-03
6.22857869e-01 1.14059344e-01 -1.97843894e-01 -1.39196682e+00
2.12022051e-01 -8.77242863e-01 6.47564113e-01 -9.74149764e-01
-1.51750457e+00 4.32300717e-01 -4.65325601e-02 -1.30553496e+00
4.30496752e-01 -5.90545833e-01 -6.56799138e-01 7.03222990e-01
-1.88298464e+00 -1.28236055e+00 -8.12310457e-01 4.95050550e-01
8.08271408e-01 -1.09883212e-01 2.77929753e-01 3.55217516e-01
-7.96554387e-01 4.83868212e-01 -7.60884285e-02 1.69403031e-01
8.53393376e-01 -1.29429388e+00 5.84997475e-01 1.08224034e+00
2.75860429e-01 4.74937588e-01 5.37452698e-01 -4.76804137e-01
-1.09820724e+00 -1.39407146e+00 9.37166989e-01 -4.35907751e-01
7.43653119e-01 -5.58105111e-01 -1.03240967e+00 7.15854943e-01
6.91390485e-02 6.12706065e-01 1.68000370e-01 1.96429402e-01
-3.17857027e-01 -3.45905162e-02 -7.60422230e-01 8.02159131e-01
1.49930954e+00 -3.59568328e-01 -7.96165645e-01 3.61793399e-01
1.06551588e+00 -3.75716567e-01 -4.81380850e-01 4.55788970e-01
2.78240293e-01 -1.03052747e+00 1.20946801e+00 -5.09475827e-01
6.00883245e-01 -5.60190201e-01 -2.09281459e-01 -1.21188319e+00
-6.92826092e-01 -5.33908367e-01 -2.21117631e-01 1.24274671e+00
1.77278191e-01 -4.13057357e-01 6.98705912e-01 1.90993652e-01
-5.09123027e-01 -8.93574774e-01 -4.93384212e-01 -7.84857929e-01
-2.12540880e-01 -3.68645370e-01 6.64557695e-01 9.93444622e-01
-2.42274582e-01 3.37172091e-01 -4.37590778e-01 1.21319741e-01
6.36726439e-01 4.88462448e-01 7.55931854e-01 -1.12035954e+00
-1.72190472e-01 -5.11569023e-01 -5.22481859e-01 -1.32367766e+00
-1.94090113e-01 -9.17797089e-01 1.23573625e-02 -1.48882473e+00
7.31955767e-01 -4.45465684e-01 -6.50809348e-01 7.21069872e-01
-7.41097391e-01 5.84054887e-01 3.89545977e-01 3.33483875e-01
-1.05048048e+00 9.04187918e-01 1.43398464e+00 -2.43641391e-01
-5.00114262e-02 -2.62535781e-01 -1.01807261e+00 6.85406804e-01
5.72796941e-01 -2.23340049e-01 -3.70532185e-01 -4.53814328e-01
6.24793954e-03 -7.28479147e-01 8.35401893e-01 -1.02228355e+00
1.56001151e-01 -2.01831266e-01 4.28571314e-01 -8.43391597e-01
2.39794329e-01 -5.84643424e-01 -2.51637340e-01 1.84679493e-01
-1.86101958e-01 -5.13958558e-02 2.37926334e-01 4.32014197e-01
-3.99229884e-01 9.82968137e-02 9.49151278e-01 8.09789822e-02
-1.40189588e+00 5.61734915e-01 2.27652431e-01 3.69123459e-01
1.12606144e+00 -9.81778875e-02 -5.75217605e-01 -1.45733848e-01
-4.47579652e-01 5.10752916e-01 6.37371838e-01 8.31092238e-01
6.40508413e-01 -1.50656283e+00 -7.23365963e-01 1.14045143e-01
6.96099997e-01 2.44664282e-01 3.78192663e-01 9.69098687e-01
-1.58549666e-01 2.93262243e-01 -2.74366260e-01 -7.09937692e-01
-1.20484829e+00 8.34233701e-01 2.37546623e-01 3.80692189e-03
-1.00984883e+00 9.62945879e-01 7.73676634e-01 -2.49556795e-01
2.96218634e-01 -1.92421839e-01 -3.74771297e-01 4.17060554e-02
7.76878357e-01 1.49002522e-01 2.20697671e-02 -6.10334098e-01
-4.85916376e-01 6.05991542e-01 -3.20872337e-01 6.30743802e-01
1.43172622e+00 -1.26835436e-01 1.15104869e-01 1.31671563e-01
1.10968673e+00 -1.36152208e-01 -1.69344640e+00 -5.93211412e-01
-1.12462886e-01 -6.22471333e-01 1.56012759e-01 -6.46854401e-01
-1.27055812e+00 6.95255101e-01 4.50420678e-01 2.32138727e-02
1.22782195e+00 3.96810770e-01 7.31597841e-01 1.32609501e-01
1.79366186e-01 -1.16980898e+00 2.92283922e-01 3.76207352e-01
1.20170164e+00 -1.55970454e+00 1.10975891e-01 -5.43851912e-01
-8.14869106e-01 6.96006596e-01 8.74521852e-01 -1.34446695e-01
4.93334651e-01 -7.58702680e-03 -6.73966669e-03 -3.46782535e-01
-6.38880968e-01 -4.16361779e-01 5.72896063e-01 5.86580694e-01
1.84884802e-01 1.12514850e-03 -1.30807176e-01 6.05636835e-01
-1.25185952e-01 -8.41770172e-02 3.75056684e-01 6.66805089e-01
-4.01865184e-01 -6.31404519e-01 -2.28675321e-01 3.42291117e-01
-3.22487563e-01 -3.03530484e-01 -3.34682375e-01 7.76170552e-01
1.29614934e-01 6.53786778e-01 1.41439661e-01 -2.12826580e-01
3.53149712e-01 -5.69113493e-01 8.29880498e-03 -6.42621636e-01
-2.39137024e-01 -9.13623646e-02 -1.02968089e-01 -7.58363545e-01
-4.57122415e-01 -9.21644926e-01 -1.05423117e+00 -1.20703489e-01
-1.18125938e-01 -7.89241865e-02 2.76997298e-01 5.81650078e-01
5.96766233e-01 4.87027824e-01 5.38518608e-01 -1.08928096e+00
-2.72290796e-01 -8.20137262e-01 -6.59801245e-01 6.95901155e-01
3.59772712e-01 -1.05441296e+00 -3.65534425e-01 -2.77980473e-02] | [9.690970420837402, -0.20385970175266266] |
c534b811-9564-4665-851f-27ff9edd703f | deep-image-orientation-angle-detection | 2007.06709 | null | https://arxiv.org/abs/2007.06709v1 | https://arxiv.org/pdf/2007.06709v1.pdf | Deep Image Orientation Angle Detection | Estimating and rectifying the orientation angle of any image is a pretty challenging task. Initial work used the hand engineering features for this purpose, where after the invention of deep learning using convolution-based neural network showed significant improvement in this problem. However, this paper shows that the combination of CNN and a custom loss function specially designed for angles lead to a state-of-the-art results. This includes the estimation of the orientation angle of any image or document at any degree (0 to 360 degree), | ['Subhadip Maji', 'Smarajit Bose'] | 2020-06-21 | null | null | null | null | ['natural-image-orientation-angle-detection'] | ['computer-vision'] | [-2.82793939e-02 8.95108357e-02 2.30702773e-01 -4.31223243e-01
-1.30946159e-01 -5.13668180e-01 4.01680946e-01 -3.40669334e-01
-4.00299519e-01 5.88710070e-01 -2.10270882e-01 -3.83235484e-01
-4.76051390e-01 -7.32282639e-01 -6.21586144e-01 -6.22532785e-01
-7.60137364e-02 2.46387944e-01 -9.64515582e-02 -4.82142031e-01
8.61782491e-01 1.27097702e+00 -1.15791392e+00 7.76957944e-02
-1.96177047e-02 1.22601759e+00 -6.61676750e-02 9.98880684e-01
7.80969039e-02 3.83612573e-01 -9.61744964e-01 -3.15341800e-01
5.25314808e-01 5.01829863e-01 -9.32472825e-01 -2.02658325e-01
9.77405548e-01 -7.19307363e-01 -4.67888951e-01 6.27162218e-01
9.34598625e-01 -2.07472831e-01 6.87812388e-01 -9.54813123e-01
-7.11737990e-01 -3.85087030e-03 -3.69579017e-01 8.95028263e-02
5.26353598e-01 -8.24168250e-02 5.50321043e-01 -7.16508865e-01
6.08382463e-01 8.41328263e-01 1.19433129e+00 -1.67046487e-03
-5.91669858e-01 -3.91277790e-01 -4.36829537e-01 3.47345233e-01
-1.60335338e+00 3.29200625e-01 8.49349856e-01 -5.21998346e-01
1.53134215e+00 1.76864415e-01 6.69145644e-01 1.10309720e+00
5.90805113e-01 3.26480031e-01 1.10406959e+00 -7.20784545e-01
-2.61428267e-01 2.10427523e-01 1.20613808e-02 5.78566432e-01
-1.37140090e-02 -7.20489584e-03 3.78440544e-02 4.53023463e-01
1.47418964e+00 1.18764840e-01 -2.82183677e-01 -5.14109060e-02
-7.37888455e-01 6.48502350e-01 5.40301323e-01 4.83700126e-01
-4.70272481e-01 1.92689434e-01 2.97956347e-01 4.99282837e-01
1.74665555e-01 6.81143999e-01 -9.50966895e-01 -2.64062285e-01
-9.91783440e-01 2.31828719e-01 7.65905678e-01 8.01349521e-01
4.58358556e-01 -3.63661982e-02 -3.42552699e-02 4.83992726e-01
1.61375061e-01 1.04135260e-01 -2.46019661e-02 -8.04489851e-01
4.15413827e-01 4.66387331e-01 7.32880384e-02 -1.18819976e+00
-7.47994125e-01 -4.24542755e-01 -8.61076593e-01 9.29365873e-01
6.31498396e-01 -5.25475681e-01 -1.37109053e+00 1.14623952e+00
5.69477081e-02 -6.90565109e-01 -3.02478969e-01 8.91493559e-01
7.12262392e-01 5.10312617e-01 -5.31346023e-01 2.85621911e-01
1.29189217e+00 -6.14240348e-01 -7.38127291e-01 2.35135585e-01
-7.06509501e-03 -1.24303615e+00 9.51800227e-01 1.08431029e+00
-7.67966568e-01 -6.20608926e-01 -1.28661144e+00 -2.77703345e-01
-8.75688612e-01 4.11160499e-01 8.92351985e-01 8.49828422e-01
-1.05724335e+00 9.96114075e-01 -3.73020828e-01 -4.31003779e-01
3.60205472e-01 8.76503885e-01 -5.42783678e-01 1.64605856e-01
-8.66364598e-01 1.22021127e+00 -2.90603824e-02 5.31956732e-01
-2.22969025e-01 -5.11591613e-01 -4.26790386e-01 9.07748416e-02
2.24596426e-01 -2.47288764e-01 1.03125691e+00 -8.65402222e-01
-1.80535829e+00 5.96457660e-01 3.63288522e-01 -5.52159995e-02
8.76353979e-01 -6.92666650e-01 -2.83849657e-01 -9.54299718e-02
-4.72172439e-01 5.56854188e-01 1.06780827e+00 -9.37643647e-01
-5.44581711e-01 -3.31068039e-01 3.83525908e-01 -1.88816860e-01
-1.82284981e-01 -6.45796731e-02 -1.31786406e-01 -7.89084196e-01
-4.58451733e-02 -7.35537291e-01 2.32241288e-01 5.82622364e-02
-5.24966002e-01 -5.70020854e-01 1.35331869e+00 -1.16187584e+00
7.26836860e-01 -1.78663552e+00 -2.38626108e-01 3.68877679e-01
-9.76547897e-02 5.30538917e-01 3.21724832e-01 4.29043949e-01
-5.73256791e-01 -4.03805170e-03 3.22738141e-01 2.64047831e-01
-2.07993016e-01 -1.79481402e-01 1.10353142e-01 5.14326751e-01
3.28213245e-01 4.82059062e-01 -2.79740185e-01 -2.45752893e-02
6.81539655e-01 1.36192632e+00 -3.70634824e-01 1.50535285e-01
3.82740378e-01 1.73273265e-01 -1.05416089e-01 4.48766917e-01
1.03923273e+00 8.83218721e-02 -1.05069175e-01 -7.30195940e-01
-3.08240443e-01 -1.40932724e-01 -1.35742271e+00 1.42047501e+00
-7.59149313e-01 1.34050286e+00 -3.03698152e-01 -9.24248993e-01
9.86768961e-01 7.19691813e-01 6.04194522e-01 -4.22764570e-01
6.30669951e-01 9.29471701e-02 7.37377256e-02 -5.59193373e-01
4.23065126e-01 2.80335367e-01 5.58562040e-01 -8.17463966e-04
2.37004682e-01 -3.72106254e-01 -3.22555870e-01 -4.22724783e-01
7.81934440e-01 2.48140886e-01 2.76415974e-01 -1.20113119e-01
6.01101935e-01 -3.69138122e-01 -4.57600728e-02 2.98237562e-01
4.73596621e-03 1.09180176e+00 5.37602007e-01 -9.96654630e-01
-1.60616326e+00 -7.31021345e-01 -2.17248127e-01 6.09580874e-01
-4.01501805e-01 5.87126464e-02 -8.75450432e-01 -6.92846000e-01
8.46830532e-02 3.80018204e-01 -7.85279036e-01 2.72534490e-01
-9.73490179e-01 -3.17758679e-01 2.44580969e-01 9.84718442e-01
7.61223674e-01 -1.06064963e+00 -5.70694566e-01 7.89650306e-02
5.16140401e-01 -9.89236534e-01 -3.58974189e-01 5.90243697e-01
-6.86820269e-01 -1.07537198e+00 -1.15673542e+00 -1.06099558e+00
6.96727872e-01 -3.84221762e-01 7.49322832e-01 -2.21014127e-01
-8.51081550e-01 1.85870662e-01 -4.08507548e-02 -5.47163248e-01
3.41640353e-01 2.43665770e-01 -4.42491591e-01 -4.70785409e-01
2.31749162e-01 -8.24265540e-01 -7.98323393e-01 1.68568149e-01
-4.90831017e-01 -4.37695146e-01 7.32270598e-01 3.96570563e-01
4.32605408e-02 -4.63441424e-02 1.85832128e-01 -3.39451909e-01
9.00163531e-01 2.87887603e-01 -6.56124413e-01 1.31719396e-01
-8.41441691e-01 1.65923417e-01 6.18706822e-01 -4.74328846e-01
-8.42692733e-01 2.78986126e-01 -4.03528750e-01 3.57469581e-02
-3.09922963e-01 -1.04935713e-01 1.11852929e-01 -4.05849963e-01
4.90546435e-01 -1.93455070e-01 -1.98745802e-01 -5.05004287e-01
2.39836752e-01 9.91016507e-01 5.37132204e-01 -3.09248529e-02
4.89643246e-01 3.19819093e-01 5.38312316e-01 -9.58034515e-01
-4.25396860e-01 -3.46792966e-01 -8.88429642e-01 -3.07563096e-01
8.37566257e-01 -3.61130267e-01 -1.18592310e+00 7.09152102e-01
-1.67762625e+00 1.00557104e-01 -3.52605641e-01 3.93523186e-01
-5.67710817e-01 2.89730221e-01 -4.53056693e-01 -6.87690437e-01
-7.00134695e-01 -1.24477065e+00 8.73607635e-01 3.41269195e-01
-2.81395644e-01 -7.73053586e-01 -3.07277851e-02 3.90011072e-02
7.01250911e-01 3.00017834e-01 6.88392520e-01 -2.65602946e-01
-4.32879090e-01 -9.56251264e-01 -6.13696516e-01 9.36481118e-01
1.42164007e-01 3.82322103e-01 -9.46774721e-01 -1.15400940e-01
-1.61565572e-01 -1.73949182e-01 2.06004158e-01 7.20416605e-01
1.42284346e+00 -1.80171624e-01 9.30441916e-02 7.93784678e-01
1.71829081e+00 2.73185939e-01 1.22936094e+00 5.75685561e-01
6.29111409e-01 1.18539453e-01 1.07110724e-01 4.62968320e-01
2.73464322e-02 7.38653123e-01 4.34793055e-01 -2.29641765e-01
-4.24475521e-01 2.66202152e-01 -3.31639260e-01 1.76435873e-01
-1.09806657e+00 -2.58526236e-01 -6.84193850e-01 3.00917149e-01
-1.16612959e+00 -7.11277306e-01 -1.79055423e-01 1.93787062e+00
5.57379007e-01 1.68254465e-01 -2.65832953e-02 8.54741633e-01
6.31647110e-01 -1.48693025e-02 -9.58152339e-02 -1.09756064e+00
1.25800269e-02 7.31125295e-01 1.09357190e+00 3.68684828e-01
-1.23197579e+00 6.58767104e-01 8.08470726e+00 6.97252870e-01
-1.79481423e+00 -2.61470169e-01 1.60860389e-01 2.94401735e-01
4.36690778e-01 -3.46283883e-01 -7.11584508e-01 1.44150704e-01
3.28462750e-01 6.52753472e-01 5.88814616e-01 1.00033247e+00
-1.27340883e-01 -4.66468409e-02 -7.95105875e-01 9.63714063e-01
1.39589101e-01 -1.01547563e+00 -2.28737384e-01 2.18739048e-01
6.88316762e-01 -3.49585861e-01 1.75298378e-01 -1.18863747e-01
-1.55732408e-01 -1.36905730e+00 2.95524269e-01 6.91793442e-01
9.57051635e-01 -9.42051351e-01 1.15132713e+00 -1.93357304e-01
-8.08659017e-01 1.06124561e-02 -2.61668861e-01 -8.52451324e-02
-7.29292957e-03 5.80037057e-01 -1.31341529e+00 3.92488688e-01
8.59866202e-01 1.22836336e-01 -3.92316312e-01 1.08125865e+00
-4.21588749e-01 2.29438275e-01 -4.82618451e-01 -1.80110276e-01
3.35919231e-01 1.88244313e-01 3.67050678e-01 1.28652883e+00
4.07908767e-01 -4.32641685e-01 -4.81149673e-01 4.88446981e-01
1.19467258e-01 3.27760763e-02 -4.85536605e-01 1.34871960e-01
-1.30460545e-01 1.38454616e+00 -4.81070876e-01 8.94714743e-02
-1.51019201e-01 1.24939680e+00 3.74606661e-02 3.45241666e-01
-5.32271028e-01 -1.32483542e+00 3.75777036e-01 3.47821265e-01
7.71619678e-01 -5.10294080e-01 -5.99473357e-01 -4.37152743e-01
3.27987075e-01 -4.18434799e-01 -1.33160114e-01 -8.86251688e-01
-8.64776909e-01 5.42820692e-01 -2.39339247e-01 -9.70200539e-01
-4.98058766e-01 -1.50661588e+00 -7.45623648e-01 1.06695008e+00
-1.22879648e+00 -1.19983053e+00 -6.04856074e-01 4.82154250e-01
4.92249846e-01 -4.82216030e-01 7.97758400e-01 3.82808954e-01
5.27735688e-02 6.68778479e-01 7.97466859e-02 4.25070912e-01
6.39825106e-01 -1.53072667e+00 1.49492517e-01 2.17707574e-01
-3.43872041e-01 5.76360345e-01 9.35617805e-01 -1.17424726e-01
-1.17249298e+00 -4.36968029e-01 8.11168075e-01 -3.35210145e-01
1.33166119e-01 -3.24521400e-02 -3.81167263e-01 7.74437308e-01
6.69464529e-01 1.37100726e-01 1.24080233e-01 1.14853807e-01
-4.44256335e-01 -2.56136209e-01 -1.67214608e+00 3.73247027e-01
7.52995729e-01 -4.83988822e-01 -5.30105233e-01 3.86672080e-01
1.52983740e-01 -4.84137028e-01 -1.09505117e+00 5.40234208e-01
1.17229390e+00 -1.12352967e+00 1.32423317e+00 -5.64639926e-01
7.39999652e-01 1.01695225e-01 7.15221390e-02 -1.13683248e+00
-3.60750616e-01 -6.01708353e-01 -5.97531535e-02 9.55260575e-01
2.26986229e-01 -6.08818471e-01 1.15684044e+00 1.02867365e-01
-2.38589980e-02 -9.97573853e-01 -8.40830445e-01 -7.52199054e-01
1.39321104e-01 -2.64278412e-01 3.88297081e-01 4.55669403e-01
-3.64101917e-01 1.41231239e-01 -3.39587659e-01 4.54809517e-02
2.54414827e-01 -4.51332927e-01 6.97932303e-01 -1.04708624e+00
1.20892711e-01 -3.88977051e-01 -1.11773944e+00 -7.73265719e-01
-3.06406438e-01 -2.66669184e-01 -7.82402903e-02 -1.74065411e+00
-3.50151658e-01 -9.51780602e-02 5.38649112e-02 3.00752550e-01
4.12491053e-01 3.37214172e-01 4.67147827e-02 -3.71575177e-01
4.10731465e-01 -2.09462419e-01 1.58098412e+00 -1.82231173e-01
7.26094171e-02 2.36102283e-01 -4.68201220e-01 8.08064103e-01
8.76996577e-01 -6.05161041e-02 -7.66320080e-02 -4.30724084e-01
3.41922224e-01 -2.74952918e-01 4.30085808e-01 -1.27620900e+00
9.06180404e-03 1.82727844e-01 9.68293309e-01 -1.04849052e+00
4.86393332e-01 -1.14736485e+00 -1.48243845e-01 3.57549816e-01
8.74441415e-02 2.60403156e-01 1.06790280e-02 -1.28530949e-01
-7.28885978e-02 -5.44431150e-01 6.84931040e-01 -1.49862695e-04
-5.62611163e-01 9.58003178e-02 -1.21953137e-01 -5.96236944e-01
8.49482954e-01 -8.04925323e-01 -2.24757016e-01 -5.94067454e-01
-8.13663483e-01 -5.43666124e-01 -5.46581112e-02 2.31243759e-01
4.82215047e-01 -1.12472153e+00 -4.36651736e-01 1.65480599e-01
-5.11597872e-01 1.81096077e-01 -1.61882781e-03 6.19900644e-01
-1.25398338e+00 7.30233192e-01 -6.71432376e-01 -6.08707309e-01
-1.38648665e+00 2.64680117e-01 4.55459476e-01 -2.69136548e-01
-5.75710714e-01 1.01362944e+00 -6.92221880e-01 -4.15823787e-01
5.25806367e-01 -4.66887981e-01 -3.42677653e-01 -1.00070819e-01
2.08367318e-01 6.84697151e-01 5.37018359e-01 -1.88161135e-01
-3.62057745e-01 1.13544035e+00 -7.86436126e-02 -3.14995125e-02
1.65656281e+00 3.63396943e-01 1.82860881e-01 -1.47463664e-01
1.77016389e+00 8.87625888e-02 -1.29539657e+00 4.23373520e-01
-3.66893142e-01 -5.07316291e-01 3.05806458e-01 -1.26434517e+00
-1.47002435e+00 1.14174771e+00 1.24878752e+00 2.27700070e-01
9.83575463e-01 -4.67128426e-01 6.36726379e-01 5.82606018e-01
4.88001443e-02 -1.42688513e+00 1.68208450e-01 8.22791100e-01
1.50601220e+00 -1.10692644e+00 3.24992508e-01 -5.14848650e-01
-1.76338494e-01 2.00712132e+00 4.46816802e-01 -6.90862119e-01
1.15485382e+00 6.38827264e-01 1.92243233e-01 -1.88630015e-01
3.09713095e-01 1.18138783e-01 4.42657799e-01 1.11777747e+00
9.11597908e-01 -1.10534556e-01 -5.06592035e-01 1.40295565e-01
-5.91626883e-01 3.10262978e-01 3.04904580e-01 1.30131614e+00
-2.42183983e-01 -9.74240839e-01 -6.23561561e-01 2.67455131e-01
-6.89994574e-01 1.56114116e-01 -4.42631662e-01 1.16434145e+00
5.67772746e-01 3.89425933e-01 2.73697317e-01 -4.22023743e-01
6.51132882e-01 -1.87197044e-01 1.22043025e+00 -1.72216445e-01
-7.71462142e-01 -3.81559730e-01 -1.08466953e-01 -2.96135634e-01
-1.05847105e-01 -2.13470429e-01 -9.72815633e-01 -4.27502483e-01
-4.17729884e-01 -5.95950067e-01 1.43053257e+00 9.21649396e-01
-4.78649698e-02 7.50969648e-01 6.73690915e-01 -1.04928637e+00
-5.68831921e-01 -1.36161697e+00 -7.06449807e-01 1.86481729e-01
4.72142279e-01 -5.71010649e-01 -2.83825278e-01 -1.40895113e-01] | [9.900273323059082, 0.06649274379014969] |
0b769ffe-d308-4f36-86ec-35777b2fada3 | a-wearable-ecg-monitor-for-deep-learning | 2201.10083 | null | https://arxiv.org/abs/2201.10083v1 | https://arxiv.org/pdf/2201.10083v1.pdf | A Wearable ECG Monitor for Deep Learning Based Real-Time Cardiovascular Disease Detection | Cardiovascular disease has become one of the most significant threats endangering human life and health. Recently, Electrocardiogram (ECG) monitoring has been transformed into remote cardiac monitoring by Holter surveillance. However, the widely used Holter can bring a great deal of discomfort and inconvenience to the individuals who carry them. We developed a new wireless ECG patch in this work and applied a deep learning framework based on the Convolutional Neural Network (CNN) and Long Short-term Memory (LSTM) models. However, we find that the models using the existing techniques are not able to differentiate two main heartbeat types (Supraventricular premature beat and Atrial fibrillation) in our newly obtained dataset, resulting in low accuracy of 58.0 %. We proposed a semi-supervised method to process the badly labelled data samples with using the confidence-level-based training. The experiment results conclude that the proposed method can approach an average accuracy of 90.2 %, i.e., 5.4 % higher than the accuracy of conventional ECG classification methods. | ['Lu Meng', 'Yang song', 'Ming Ding', 'Zijiao Chen', 'Xucun Yan', 'Zihuai Lin', 'Peng Wang'] | 2022-01-25 | null | null | null | null | ['ecg-classification'] | ['medical'] | [ 1.93287984e-01 -9.45805460e-02 2.05639422e-01 -3.79229724e-01
-5.54233432e-01 -2.42634997e-01 -9.30381939e-02 3.02516311e-01
-5.17760336e-01 9.43110049e-01 -3.17839056e-01 -5.46623051e-01
-9.13194418e-02 -8.57140839e-01 -2.17423871e-01 -6.80097818e-01
-3.47795069e-01 6.64372370e-02 -2.73993194e-01 1.78020313e-01
1.51908904e-01 3.87085289e-01 -1.01809049e+00 2.07666248e-01
8.85800719e-01 1.33091414e+00 -4.46047664e-01 7.88779974e-01
1.32437915e-01 7.81179845e-01 -1.03760970e+00 -3.34319696e-02
7.06967190e-02 -8.89439523e-01 -5.39057970e-01 -4.48620766e-01
-1.31017059e-01 -4.03246254e-01 1.40330195e-01 7.64035046e-01
1.15343499e+00 -3.95671755e-01 3.11308503e-01 -7.50416279e-01
-3.88419658e-01 5.13469577e-01 -5.12787461e-01 3.94298673e-01
7.32319579e-02 -8.21836367e-02 7.13761300e-02 -3.68758917e-01
1.30360248e-02 5.63427210e-01 1.26257122e+00 5.12292206e-01
-6.58128917e-01 -8.87967348e-01 -3.99899572e-01 1.43292472e-01
-1.44953775e+00 -2.10507870e-01 8.18692327e-01 -2.82529086e-01
6.02510333e-01 2.59266973e-01 1.01961040e+00 1.01377952e+00
7.65084028e-01 1.92162797e-01 1.20909715e+00 -3.18946093e-01
3.72017995e-02 3.28653045e-02 1.39016122e-01 4.81941849e-01
3.69227260e-01 2.32660621e-01 -7.03971088e-02 -3.89243692e-01
9.15850401e-01 2.95789123e-01 -2.38081440e-01 3.46223921e-01
-1.13676989e+00 5.07388771e-01 4.31071639e-01 5.84469914e-01
-4.52547491e-01 -3.50589603e-02 7.34781325e-01 3.79568309e-01
2.05393419e-01 2.70825326e-01 -4.92812663e-01 -3.50500703e-01
-1.04714179e+00 -9.52513441e-02 7.32949793e-01 4.42995340e-01
1.33402050e-01 2.40239620e-01 -1.63327977e-01 5.52333534e-01
2.85615265e-01 4.79132414e-01 5.89956760e-01 -5.95231354e-01
1.76440999e-01 3.89558256e-01 9.38833132e-02 -1.23311877e+00
-6.47493899e-01 -1.06037998e+00 -1.53662252e+00 5.31569235e-02
2.67507821e-01 -7.06664026e-01 -7.39106715e-01 1.40390980e+00
-2.16160789e-02 3.11460376e-01 1.29054427e-01 7.83407032e-01
9.74610031e-01 5.11517763e-01 3.29285890e-01 -5.61229885e-01
1.16198957e+00 -4.89188492e-01 -1.07024586e+00 9.72012058e-02
5.96514523e-01 -4.90474910e-01 5.74922681e-01 6.73622191e-01
-8.45347226e-01 -7.81333029e-01 -1.41383421e+00 4.38099295e-01
-9.66331735e-02 1.32918581e-01 4.07121062e-01 1.10722184e+00
-8.03533494e-01 9.64392304e-01 -7.35394716e-01 -2.63220578e-01
5.35537779e-01 1.58146232e-01 -3.42802890e-02 3.31612676e-01
-1.55658019e+00 7.77455926e-01 2.50181824e-01 6.85858548e-01
-7.59009719e-01 -2.22417757e-01 -5.09321690e-01 -2.34467294e-02
-2.13028967e-01 -6.06915653e-01 8.54871154e-01 -9.30310607e-01
-1.43439758e+00 6.59067929e-01 2.04335436e-01 -6.36077762e-01
6.69182897e-01 -3.94780874e-01 -7.09733963e-01 1.10227220e-01
-1.88857079e-01 8.02379772e-02 7.50317752e-01 -7.31928229e-01
-4.02187228e-01 -4.59098876e-01 -3.38600874e-01 -3.73321734e-02
-3.66373032e-01 1.33630270e-02 1.47142321e-01 -6.10462427e-01
1.07023925e-01 -6.75915182e-01 -1.55721799e-01 -2.38701552e-01
-2.09241211e-01 7.06277639e-02 6.15025401e-01 -9.80912864e-01
1.48911691e+00 -2.07709050e+00 -4.82309580e-01 2.33237550e-01
3.86438519e-01 6.95492506e-01 4.74522084e-01 2.77303427e-01
-2.70737521e-02 4.79851007e-01 -4.24959958e-01 4.20858003e-02
-5.86980045e-01 2.85760891e-02 8.99042040e-02 4.80939388e-01
-2.00069606e-01 7.72423983e-01 -7.26845384e-01 -5.20983517e-01
2.29634121e-01 6.07265353e-01 1.15895838e-01 3.69813412e-01
6.05612516e-01 8.37262392e-01 -3.66525471e-01 5.57833433e-01
6.80738688e-01 -1.03934854e-01 1.74384698e-01 -9.23688337e-03
1.31510869e-01 -4.52376120e-02 -8.62580001e-01 1.76843798e+00
-3.05588245e-01 2.65006125e-01 -3.44961435e-01 -1.02192461e+00
1.31595814e+00 8.07107985e-01 5.24957657e-01 -6.03340983e-01
4.36970055e-01 3.06722194e-01 1.87615708e-01 -7.87754774e-01
-3.81285459e-01 -4.26570296e-01 3.38985398e-02 3.33308101e-01
-2.27640212e-01 5.62720299e-01 -5.19705415e-01 -3.35354298e-01
1.07585180e+00 3.39739509e-02 3.93336117e-01 -3.93706620e-01
5.39532423e-01 -4.69346941e-01 9.63046134e-01 8.75383139e-01
-5.62076092e-01 6.88863456e-01 2.31725574e-01 -1.06141400e+00
-5.32129943e-01 -7.17733026e-01 -3.16988677e-01 1.01782642e-01
-4.96302135e-02 -2.34373480e-01 -7.81723678e-01 -6.22136474e-01
-3.10049564e-01 1.65225998e-01 -3.66766632e-01 -4.48241681e-01
-5.48378706e-01 -9.99050975e-01 1.23101377e+00 7.21507132e-01
1.03761351e+00 -1.23640788e+00 -1.27968955e+00 4.97296721e-01
-2.32167825e-01 -6.81690931e-01 1.40677705e-01 1.40851110e-01
-1.18364990e+00 -9.28565741e-01 -9.66025591e-01 -7.19390154e-01
2.19136968e-01 -5.61667979e-01 1.02406728e+00 3.93648744e-01
-3.88614744e-01 -3.72980177e-01 -3.85394007e-01 -6.97076201e-01
-2.43421510e-01 9.95135009e-02 -3.10648885e-02 2.35609412e-01
4.18624371e-01 -7.15778351e-01 -9.88445699e-01 4.22923593e-03
-4.06846136e-01 -1.47817537e-01 6.09731793e-01 6.62875056e-01
3.70242089e-01 7.50510693e-02 1.34119487e+00 -1.06107819e+00
8.04975033e-01 -3.64729315e-01 -1.42820418e-01 6.43207207e-02
-1.04612172e+00 -4.32547569e-01 7.50180483e-01 -2.33709767e-01
-6.49045885e-01 -9.83122811e-02 -5.60941339e-01 -1.85667187e-01
-5.53114235e-01 6.93837464e-01 -2.39166021e-02 -1.68547053e-02
6.17832422e-01 1.55308977e-01 -1.47567555e-01 -4.47996408e-01
-4.80750650e-01 1.00021291e+00 4.79823321e-01 -1.87553748e-01
2.57451415e-01 1.70121446e-01 3.74458134e-02 -6.31737292e-01
-6.92739487e-01 -1.49295017e-01 -4.86760765e-01 -3.65177691e-01
1.00476968e+00 -8.12940836e-01 -7.37960875e-01 7.85251141e-01
-1.17213702e+00 -3.88757326e-04 -9.36475098e-02 5.45285821e-01
-1.03112869e-01 5.54129303e-01 -8.56622934e-01 -1.07205009e+00
-9.18650627e-01 -6.42008722e-01 6.53019071e-01 6.43151328e-02
-2.92336673e-01 -8.37609470e-01 1.33952601e-02 5.09352945e-02
7.50547945e-01 1.02088547e+00 8.15980792e-01 -5.07005930e-01
1.95022181e-01 -6.01955056e-01 7.03515932e-02 6.21622264e-01
2.21977323e-01 -3.24027807e-01 -1.25289452e+00 -3.17055732e-01
6.09824657e-01 -1.29843771e-01 6.68184996e-01 4.80500758e-01
1.37719500e+00 -3.81620489e-02 -8.89861286e-02 7.37566888e-01
1.29092944e+00 8.44352245e-01 1.03466713e+00 1.71772450e-01
5.76494455e-01 -2.49533225e-02 4.12640095e-01 4.62493747e-01
5.71845733e-02 7.03438744e-03 1.40017301e-01 -5.84245682e-01
3.14207315e-01 -3.54431532e-02 6.99596778e-02 8.61692667e-01
-3.86984378e-01 -4.43700105e-02 -1.08673096e+00 3.17812532e-01
-1.82933533e+00 -7.98622906e-01 -3.87644410e-01 2.38728523e+00
7.33780384e-01 2.64985800e-01 1.29034743e-02 6.98063612e-01
8.94651711e-01 -1.44607291e-01 -4.96895403e-01 -5.88790298e-01
1.51496693e-01 4.85931456e-01 2.85815239e-01 7.40120634e-02
-1.22366726e+00 1.94493949e-01 6.35838270e+00 2.54755855e-01
-1.42966568e+00 2.40733832e-01 9.72188175e-01 3.71464998e-01
3.47335666e-01 -2.26206318e-01 -3.24334383e-01 7.28033841e-01
1.38303721e+00 1.46664932e-01 -1.72095627e-01 6.01651490e-01
2.60953397e-01 2.38529116e-01 -9.01396334e-01 1.48025250e+00
1.23281442e-01 -9.48588014e-01 -1.16155669e-01 -1.94187224e-01
2.25458190e-01 -4.17328328e-01 -2.90386945e-01 1.92969039e-01
-8.05536509e-01 -1.21951771e+00 2.97149211e-01 7.92519689e-01
1.05905831e+00 -8.88800144e-01 1.49809492e+00 5.47472715e-01
-9.70371604e-01 -1.29070908e-01 -2.33389124e-01 -5.53259969e-01
7.86021799e-02 9.24170911e-01 -3.67352456e-01 6.54488802e-01
8.80034983e-01 7.98813224e-01 -4.53286350e-01 1.06690681e+00
-4.27466333e-02 9.71168280e-01 -1.68010205e-01 5.43342158e-02
-1.88300475e-01 1.30205033e-02 2.74449944e-01 8.86749983e-01
3.89009476e-01 9.04971808e-02 2.06000686e-01 6.22098386e-01
3.25152054e-02 1.59104854e-01 -7.11391747e-01 2.32576862e-01
1.14851117e-01 1.16100824e+00 -7.27835417e-01 -5.87095201e-01
-1.93859085e-01 8.26120973e-01 -3.11288625e-01 1.61801770e-01
-1.01689482e+00 -9.16097462e-01 -3.21306497e-01 2.37765849e-01
-3.59042853e-01 2.11167619e-01 -7.09803879e-01 -1.02658415e+00
1.90949082e-01 -7.38683999e-01 4.63589787e-01 -3.77053052e-01
-1.20839453e+00 9.00281429e-01 -4.41445768e-01 -1.32251298e+00
-9.10286382e-02 -1.77402914e-01 -8.40678990e-01 1.06321836e+00
-1.30970252e+00 -7.86719203e-01 -5.72614789e-01 5.63749433e-01
1.50415301e-01 -2.69834865e-02 1.38652599e+00 7.45064974e-01
-5.17569900e-01 6.33950174e-01 -3.49801779e-01 4.59956527e-01
7.30969846e-01 -1.04294622e+00 7.60460719e-02 7.82122135e-01
-3.20457816e-01 6.78952456e-01 2.37541676e-01 -5.20747781e-01
-8.04294705e-01 -1.16211295e+00 1.08043528e+00 -1.20100774e-01
-2.35036224e-01 -1.19422013e-02 -8.06852818e-01 4.20352727e-01
3.86404723e-01 1.67597398e-01 8.87766302e-01 1.83805916e-02
2.38589257e-01 -5.50720930e-01 -1.33582747e+00 -1.97930057e-02
6.63734734e-01 -2.96206355e-01 -5.55557072e-01 -3.95993479e-02
1.29922032e-01 -1.50282964e-01 -1.13323486e+00 9.27703738e-01
1.00006056e+00 -1.12004745e+00 6.39742434e-01 -4.59275633e-01
1.65446579e-01 -2.30618551e-01 3.94318104e-01 -1.05014706e+00
-3.37855339e-01 -6.04366481e-01 -1.25376359e-01 1.13021004e+00
3.31500649e-01 -8.87330890e-01 5.31678259e-01 3.48485798e-01
-1.76903065e-02 -9.01125312e-01 -6.22662365e-01 -5.65127492e-01
-4.04793397e-03 -2.40387812e-01 4.76993799e-01 1.11014223e+00
1.79889053e-02 4.84674901e-01 -6.37808144e-01 -8.08217470e-03
6.70184255e-01 3.11055891e-02 2.63045520e-01 -1.56313968e+00
-2.01276556e-01 1.89964831e-01 -5.62573016e-01 -4.38579679e-01
-4.55916911e-01 -6.51524961e-01 -1.46728799e-01 -1.47477460e+00
7.95845538e-02 -6.45735681e-01 -1.12452519e+00 4.62043643e-01
-1.83561876e-01 4.20953810e-01 -2.78706342e-01 2.09730789e-02
-2.77421594e-01 1.51989341e-01 1.07764244e+00 -2.26401538e-03
-3.41300964e-01 3.25404674e-01 -7.17766285e-01 9.42747235e-01
1.37217319e+00 -5.37329495e-01 -3.35401416e-01 -4.26753730e-01
2.37399206e-01 5.21958709e-01 2.90684134e-01 -1.63769352e+00
4.23247814e-02 3.25240612e-01 9.93068874e-01 -3.88858825e-01
-9.26937535e-02 -8.05771887e-01 2.99835563e-01 9.48771715e-01
-3.38078707e-01 2.52705783e-01 7.51740932e-02 4.32472020e-01
-2.90202111e-01 8.54071975e-02 8.00701141e-01 -3.32826018e-01
-1.09327108e-01 2.54785687e-01 -5.34453452e-01 -1.28246590e-01
1.01047897e+00 -1.41719282e-01 1.10554956e-01 -3.18429053e-01
-9.72333729e-01 2.79391315e-02 -1.86847597e-01 2.35173851e-01
9.33611691e-01 -1.26191473e+00 -7.27717400e-01 3.91206026e-01
1.32300751e-02 -1.48295924e-01 4.65068221e-01 1.13625848e+00
-8.53878200e-01 1.35196671e-01 -4.52893168e-01 -7.16063499e-01
-1.20348358e+00 3.50693136e-01 6.17480397e-01 -2.15122342e-01
-1.00532520e+00 3.99437577e-01 -4.62330312e-01 -6.92810044e-02
4.67181891e-01 -5.50359011e-01 -5.44621050e-01 -2.16303930e-01
5.58145344e-01 4.60453123e-01 2.76016384e-01 -1.65794268e-01
-5.22173941e-01 6.25473142e-01 3.26661170e-01 2.09808528e-01
1.08013940e+00 5.32058738e-02 -1.67292118e-01 7.04925835e-01
9.43006456e-01 -3.11924130e-01 -4.68398541e-01 4.17879552e-01
-1.66661739e-01 -2.16299444e-01 1.06351614e-01 -1.08332288e+00
-1.10966516e+00 1.37494981e+00 1.42021632e+00 3.20823222e-01
1.35008383e+00 -7.46125400e-01 1.29629159e+00 4.30743933e-01
5.18507481e-01 -8.69384170e-01 -4.82589185e-01 -1.97436828e-02
4.46503490e-01 -1.01823711e+00 -2.13928297e-01 7.67438337e-02
-3.77556235e-01 1.23342156e+00 3.02042812e-01 -3.18303138e-01
1.07220316e+00 2.49174476e-01 6.20313644e-01 -1.26470849e-01
-3.69491518e-01 3.95664245e-01 -1.95596054e-01 5.32510340e-01
6.76559031e-01 1.58867031e-01 -7.74237871e-01 9.83576357e-01
1.75016224e-01 5.46058357e-01 3.63669008e-01 1.08036351e+00
-2.46424928e-01 -1.02629685e+00 -2.39123926e-01 7.65394449e-01
-1.12093425e+00 -2.33860090e-02 -2.67391860e-01 2.21037403e-01
5.39144456e-01 1.13841569e+00 -2.18283921e-01 -6.29666030e-01
8.23184699e-02 4.12235707e-01 2.05183700e-01 -4.56804633e-01
-8.23574960e-01 8.31758156e-02 1.95087530e-02 -2.52910852e-01
-6.44896984e-01 -1.10227495e-01 -1.14121175e+00 -4.84086089e-02
-2.78929293e-01 3.19951564e-01 4.15149838e-01 8.38717759e-01
4.42760229e-01 7.86107481e-01 5.72483897e-01 -3.33889723e-01
-3.71243417e-01 -1.27702963e+00 -8.26681793e-01 3.24728519e-01
4.68574405e-01 -2.74372578e-01 -2.15621024e-01 1.69114456e-01] | [14.274114608764648, 3.2542214393615723] |
0f21e511-3d45-44b6-a889-c3a64fefbe07 | neuromorphic-computing-for-content-based | 2008.01380 | null | https://arxiv.org/abs/2008.01380v2 | https://arxiv.org/pdf/2008.01380v2.pdf | Neuromorphic Computing for Content-based Image Retrieval | Neuromorphic computing mimics the neural activity of the brain through emulating spiking neural networks. In numerous machine learning tasks, neuromorphic chips are expected to provide superior solutions in terms of cost and power efficiency. Here, we explore the application of Loihi, a neuromorphic computing chip developed by Intel, for the computer vision task of image retrieval. We evaluated the functionalities and the performance metrics that are critical in content-based visual search and recommender systems using deep-learning embeddings. Our results show that the neuromorphic solution is about 2.5 times more energy-efficient compared with an ARM Cortex-A72 CPU and 12.5 times more energy-efficient compared with NVIDIA T4 GPU for inference by a lightweight convolutional neural network without batching while maintaining the same level of matching accuracy. The study validates the potential of neuromorphic computing in low-power image retrieval, as a complementary paradigm to the existing von Neumann architectures. | ['Te-Yuan Liu', 'Daniel Prusinski', 'Luis Stevens', 'Ata Mahjoubfar'] | 2020-08-04 | null | null | null | null | ['content-based-image-retrieval'] | ['computer-vision'] | [ 1.89998522e-01 -4.13048834e-01 2.70381153e-01 -6.46125153e-02
1.63725406e-01 -3.75456840e-01 5.44517696e-01 1.21906232e-02
-9.74253237e-01 3.35791677e-01 -2.73507833e-01 -3.24260175e-01
-6.97486475e-02 -9.41837728e-01 -7.02498794e-01 -5.92216313e-01
-1.30789235e-01 1.49675146e-01 1.94564775e-01 3.89305130e-02
3.84113520e-01 6.95121109e-01 -2.26223302e+00 2.76211530e-01
1.21392146e-01 1.67285061e+00 3.69692266e-01 5.20715117e-01
-8.14053342e-02 5.20496726e-01 -6.33985043e-01 -1.86254576e-01
3.46533269e-01 3.13467830e-01 -2.88529128e-01 -9.13413823e-01
3.31662923e-01 -2.75652736e-01 -1.02500975e+00 9.81048763e-01
6.82111621e-01 -2.69690573e-01 4.98693049e-01 -1.11491072e+00
-4.41482097e-01 2.30323464e-01 5.34883849e-02 5.45890152e-01
-4.12790067e-02 1.71831205e-01 4.59649116e-01 -7.36418724e-01
2.75578678e-01 9.32771862e-01 5.93878984e-01 6.79825366e-01
-8.86874020e-01 -8.92433345e-01 -7.46960759e-01 4.58576888e-01
-1.41037714e+00 -6.15161598e-01 2.44896665e-01 1.52311474e-01
1.89502573e+00 1.50251985e-01 1.27857530e+00 9.50628102e-01
1.17768180e+00 2.03581527e-01 1.08597302e+00 -1.89199045e-01
8.56192231e-01 -2.20596835e-01 4.73301381e-01 7.60031760e-01
7.00350761e-01 1.88832954e-01 -1.10963142e+00 -3.67259771e-01
3.42713267e-01 2.16950715e-01 9.79755595e-02 2.58908302e-01
-7.84350276e-01 3.18374604e-01 4.81788784e-01 3.58299673e-01
-2.51435727e-01 1.37806523e+00 7.60864854e-01 2.67366767e-01
-1.46899253e-01 3.85747313e-01 -6.81708902e-02 -2.02352762e-01
-9.12030816e-01 1.68731257e-01 1.06664169e+00 8.63667727e-01
6.52691722e-01 4.22560543e-01 -1.02346957e-01 5.26202559e-01
3.49345744e-01 8.91883612e-01 9.44094062e-01 -1.21443975e+00
-6.11506820e-01 6.92570448e-01 -5.37086427e-01 -4.24368113e-01
-3.52763563e-01 -2.85805106e-01 -8.13479006e-01 3.40629667e-01
-1.22676261e-01 3.25521529e-01 -9.02189612e-01 1.40561008e+00
-2.61144519e-01 4.84925620e-02 5.73836677e-02 8.88796687e-01
8.37619603e-01 6.98750496e-01 7.83309713e-02 3.32005888e-01
1.73021030e+00 -6.23784602e-01 -5.64197540e-01 -1.79677784e-01
1.76065177e-01 -5.43955684e-01 5.51776171e-01 2.76559919e-01
-1.09410918e+00 -4.88428473e-01 -1.75775671e+00 -2.90495247e-01
-8.27723980e-01 -1.69920340e-01 9.47523057e-01 9.30527866e-01
-1.54490578e+00 6.58175588e-01 -8.98902059e-01 -4.08799917e-01
4.75823581e-01 7.72860289e-01 7.83532709e-02 1.13383882e-01
-8.80078256e-01 7.81382203e-01 3.66793275e-01 -2.87331223e-01
-9.54222798e-01 -8.77299070e-01 -2.77570039e-01 6.52363658e-01
-6.18535221e-01 -7.59617925e-01 1.10347319e+00 -5.97974420e-01
-1.45490444e+00 1.00049233e+00 -4.61168438e-02 -8.85275066e-01
-2.37048641e-01 3.59602660e-01 -4.46699560e-01 2.67130673e-01
-3.59420121e-01 1.09359396e+00 6.43564820e-01 -2.49657124e-01
-2.34082595e-01 -7.86105394e-01 -2.48143479e-01 -4.66365725e-01
-9.75922942e-01 -1.66531682e-01 -4.22730505e-01 -8.70819911e-02
-2.05203235e-01 -1.11454844e+00 7.84261301e-02 5.55624008e-01
2.30744764e-01 -2.65305012e-01 1.20983577e+00 2.84897666e-02
7.38387823e-01 -2.10105610e+00 -5.29135466e-01 3.86090249e-01
2.82622665e-01 5.70824087e-01 -1.04756080e-01 1.90843180e-01
3.55530113e-01 -2.46501774e-01 1.27999529e-01 1.29547909e-01
1.68552063e-02 3.07169318e-01 -3.41277748e-01 4.39859420e-01
1.13482259e-01 1.12968791e+00 -4.98198539e-01 -3.69575731e-02
6.30862638e-02 7.20570683e-01 -4.33313906e-01 1.77195226e-03
-5.43570705e-02 -3.77723038e-01 -8.61085355e-02 8.13880622e-01
5.88984370e-01 -2.95506120e-01 3.87900174e-01 -3.53974849e-01
-3.01931292e-01 6.91985786e-02 -4.60940182e-01 2.11296225e+00
-5.20472407e-01 1.18225420e+00 1.66378487e-02 -6.77085042e-01
1.03643942e+00 -9.14111435e-02 3.03378284e-01 -1.68863738e+00
4.84424740e-01 5.86064994e-01 1.26673326e-01 -1.98148876e-01
5.19607484e-01 3.87888968e-01 -4.53023165e-02 5.29281676e-01
3.90796185e-01 1.51728645e-01 -1.45815534e-03 -1.50013175e-02
1.88593411e+00 -1.32245854e-01 -2.67074734e-01 -1.06331241e+00
1.52750220e-02 -2.20925864e-02 1.54673651e-01 7.90940046e-01
-2.52498925e-01 4.96497862e-02 1.78137854e-01 -7.47831464e-01
-1.33079886e+00 -1.21252251e+00 -6.18372679e-01 8.23914886e-01
1.52736589e-01 -4.68853772e-01 -9.63447332e-01 6.92962468e-01
1.04003578e-01 2.18193516e-01 -2.71082968e-01 -4.86691773e-01
-3.34722310e-01 -8.51456761e-01 1.19840598e+00 4.46672887e-01
7.71592379e-01 -9.49699879e-01 -2.07731986e+00 2.73784965e-01
6.70353889e-01 -9.67479706e-01 5.22240773e-02 1.00389946e+00
-8.74332488e-01 -6.44691408e-01 -3.43328178e-01 -7.26947308e-01
3.36820424e-01 2.36567095e-01 1.08935905e+00 2.78798550e-01
-1.34607613e+00 4.04951543e-01 1.45760655e-01 -5.49298227e-01
1.77456602e-03 1.55444475e-04 2.84854412e-01 -6.11436069e-01
8.14676881e-01 -9.39293385e-01 -1.24751103e+00 -3.09727013e-01
-1.08111739e+00 -3.64912897e-01 6.20007396e-01 7.99315572e-01
6.92784727e-01 -2.94253767e-01 2.80390918e-01 -3.55912000e-01
5.88188469e-01 -4.24900472e-01 -7.10388184e-01 1.31141189e-02
-1.06030452e+00 4.29323316e-01 8.46585631e-01 -3.60975713e-01
-4.34383363e-01 2.10962504e-01 -3.99173647e-02 -4.86215800e-01
4.18974966e-01 1.04285471e-01 2.28728801e-01 -6.48972452e-01
7.24555969e-01 3.57564390e-01 4.69869152e-02 -8.28526542e-02
-1.97762161e-01 7.14256763e-01 6.70102000e-01 -3.81659329e-01
-9.26885828e-02 6.43966138e-01 4.95035142e-01 -8.39797258e-01
3.82982284e-01 -1.17713444e-01 1.35573328e-01 -4.44249004e-01
6.35655940e-01 -9.11251783e-01 -1.34621549e+00 5.66128910e-01
-1.32571709e+00 -1.20427720e-01 -1.25107586e-01 3.31289656e-02
-3.66216123e-01 -4.33815941e-02 -7.87505031e-01 -6.78317428e-01
-1.27590322e+00 -8.36388290e-01 8.51452470e-01 7.21030295e-01
-1.12222701e-01 -4.10335958e-01 2.97215164e-01 -1.14786156e-01
9.33268309e-01 -2.36962542e-01 1.26537740e+00 -4.45134580e-01
-7.28438199e-01 -5.01064993e-02 -4.72214639e-01 4.05953676e-02
-6.16974711e-01 -1.28871202e-01 -1.55726314e+00 -4.62645352e-01
-9.24750194e-02 -5.39452732e-01 1.16169822e+00 1.50612384e-01
1.53749096e+00 2.84950715e-02 -3.45412970e-01 8.90387297e-01
2.05222487e+00 4.42972839e-01 8.00503910e-01 2.07125843e-01
-4.28433120e-02 2.95406580e-01 -3.90934408e-01 6.10943258e-01
-9.62302163e-02 3.76375079e-01 6.34149671e-01 3.76016587e-01
-2.81310707e-01 -4.44204137e-02 7.26050958e-02 1.03521311e+00
4.03121203e-01 -1.36674836e-01 -1.02045250e+00 4.81771171e-01
-1.60541677e+00 -6.06672347e-01 2.55456001e-01 2.08371496e+00
4.37761664e-01 -3.93605009e-02 -4.31337714e-01 4.04441394e-02
3.17665398e-01 -3.03946137e-01 -1.17474818e+00 -1.02609754e+00
-2.22737685e-01 1.21127248e+00 8.35985065e-01 -3.84528875e-01
-3.10133725e-01 4.42038268e-01 6.52659702e+00 8.22447896e-01
-1.36159432e+00 5.08856356e-01 3.57112646e-01 -7.20397532e-01
-2.58245260e-01 -2.15449482e-01 -4.00226027e-01 8.28745902e-01
1.82264447e+00 -2.01393291e-01 7.14808762e-01 8.38438869e-01
-3.41135263e-01 -3.84761393e-01 -1.24673045e+00 1.42840624e+00
-4.77507301e-02 -1.79625714e+00 5.48745915e-02 3.03276509e-01
3.81090641e-01 4.19943064e-01 2.75692642e-01 1.78908452e-01
-3.91712725e-01 -1.21718359e+00 6.91827178e-01 6.29266143e-01
9.19509232e-01 -9.41923559e-01 6.26590431e-01 -4.62148339e-03
-9.98512566e-01 -2.89852351e-01 -9.36550379e-01 -3.28401089e-01
-5.68729937e-01 5.06175280e-01 -1.87740952e-01 -3.35873753e-01
1.47689009e+00 2.42603160e-02 -4.55612928e-01 1.05747414e+00
7.81333923e-01 1.91299155e-01 -5.04966557e-01 -9.06816542e-01
7.03644529e-02 -1.62199866e-02 3.04483455e-02 1.41773140e+00
5.56727290e-01 6.01140130e-03 -6.79949939e-01 1.14382613e+00
-5.23039401e-01 -2.48377845e-01 -8.90742779e-01 -2.61723489e-01
8.05695057e-01 1.44595861e+00 -7.59022892e-01 -3.96428704e-01
-1.00824021e-01 1.06630325e+00 2.12814271e-01 -1.75442681e-01
-5.18825948e-01 -8.25716734e-01 9.08991039e-01 4.37342264e-02
7.59869963e-02 -2.25153476e-01 -5.87062180e-01 -6.25086010e-01
6.79159816e-03 -3.05785269e-01 -3.16167295e-01 -9.75302339e-01
-7.91766405e-01 6.97239637e-01 -5.21493137e-01 -7.05424130e-01
-2.38583520e-01 -1.27634490e+00 -6.56401277e-01 5.88556707e-01
-1.33862317e+00 -5.93709648e-01 -5.19456983e-01 3.86322677e-01
4.13776413e-02 -4.18064326e-01 1.36677420e+00 5.45329332e-01
-3.32671016e-01 6.46142840e-01 5.91457963e-01 -3.50488782e-01
1.21034622e-01 -6.06293738e-01 4.81223404e-01 3.84820491e-01
7.25506172e-02 7.37814069e-01 4.02803212e-01 -7.37573355e-02
-2.56396055e+00 -1.02434850e+00 4.72470939e-01 3.35267603e-01
5.92451692e-01 -5.79180479e-01 -7.97571301e-01 -1.73870012e-01
5.91916740e-01 1.50877489e-02 8.03082228e-01 -6.40589535e-01
-5.03790677e-01 -5.67249477e-01 -1.31807446e+00 6.37325704e-01
1.17982554e+00 -8.70158434e-01 -6.28411248e-02 2.28562623e-01
2.33109370e-01 2.13572443e-01 -7.76736498e-01 2.31600091e-01
1.08296144e+00 -9.22889709e-01 8.11840892e-01 -7.01777190e-02
3.43187779e-01 3.90831493e-02 -3.66111189e-01 -7.57731736e-01
-2.80963242e-01 -2.33354330e-01 -2.64461607e-01 6.29775584e-01
5.82234643e-04 -7.53921211e-01 1.04388463e+00 5.64541817e-01
-4.40069363e-02 -8.47134531e-01 -1.63003910e+00 -8.22340429e-01
-1.17245868e-01 -3.62637043e-01 3.53322864e-01 2.12423299e-02
9.51575041e-02 3.70067023e-02 1.77056819e-01 -3.92775059e-01
7.61387467e-01 3.95709686e-02 6.12556338e-02 -1.15815592e+00
-2.29796022e-01 -5.99436164e-01 -1.18324411e+00 -6.25007331e-01
3.65310907e-01 -1.05621099e+00 -4.70646694e-02 -1.00928640e+00
5.31030715e-01 -1.49789929e-01 -6.24629557e-01 4.31933224e-01
7.72682488e-01 8.05833757e-01 1.12792544e-01 2.33097360e-01
-3.94537002e-01 2.62628883e-01 4.22782332e-01 -4.91582990e-01
5.63786626e-01 -7.89310396e-01 -2.69389689e-01 2.70595968e-01
7.13484764e-01 -4.46425259e-01 -4.18776095e-01 -7.43807614e-01
2.46480137e-01 -2.46917650e-01 4.69397128e-01 -1.77288783e+00
1.00802267e+00 6.74985588e-01 7.00552046e-01 -2.52654046e-01
7.93910325e-01 -8.30958903e-01 5.73504090e-01 1.26865768e+00
-1.65985823e-02 4.52609450e-01 5.15744328e-01 4.88100260e-01
1.31222531e-01 -5.35051882e-01 6.08395159e-01 -9.71694291e-02
-1.04178298e+00 1.94069877e-01 -1.21634459e+00 -4.01388139e-01
8.72636378e-01 -5.34362972e-01 -8.25554073e-01 3.16974640e-01
3.62667851e-02 -2.26191908e-01 4.74937737e-01 1.49823472e-01
9.77597475e-01 -1.21421623e+00 -1.21423751e-01 4.63558346e-01
1.15477458e-01 -9.32592988e-01 8.72600228e-02 1.52879730e-01
-9.93813038e-01 7.30039656e-01 -1.13223183e+00 -6.32405400e-01
-9.73600328e-01 3.21082115e-01 3.91216427e-01 2.25023106e-01
-3.16670239e-01 5.96368194e-01 -4.95317698e-01 1.84843510e-01
3.50102574e-01 -1.13377489e-01 4.25635815e-01 -3.86725575e-01
5.31719506e-01 5.68740249e-01 5.00228047e-01 1.84586257e-01
-7.63919771e-01 4.32919204e-01 1.18942156e-01 -1.70614168e-01
1.13334036e+00 5.19125521e-01 -6.18541002e-01 2.38726094e-01
1.47316241e+00 -8.05904090e-01 -6.22188151e-01 7.51909167e-02
7.19973147e-02 4.91038077e-02 4.97582167e-01 -7.32826710e-01
-1.14452636e+00 9.20026600e-01 1.51880693e+00 -1.49049655e-01
1.26291335e+00 -3.70972484e-01 1.03533256e+00 1.13832080e+00
8.76733661e-01 -1.49589634e+00 1.67930454e-01 5.75062037e-01
2.87742525e-01 -5.53268850e-01 -1.29288107e-01 2.74857432e-01
5.05510032e-01 1.45035279e+00 6.76490605e-01 -6.02105141e-01
8.76624882e-01 1.22316742e+00 -3.34710985e-01 -3.08977932e-01
-1.20771599e+00 7.32682124e-02 -1.92597464e-01 4.09634620e-01
3.08983386e-01 1.24833047e-01 -3.93342376e-01 2.13015631e-01
-7.91344345e-02 2.88844198e-01 2.62667686e-01 1.18598473e+00
-4.52841252e-01 -8.41985822e-01 9.61851776e-02 8.08625758e-01
-4.63999480e-01 -2.67564178e-01 -3.15109164e-01 5.24493344e-02
4.43606377e-02 7.03616261e-01 8.62946391e-01 -8.54015648e-01
6.35468364e-02 7.59632736e-02 8.02026451e-01 -1.12527788e-01
-1.19529128e+00 -5.45867980e-01 -2.23446667e-01 -7.64205098e-01
1.33415256e-02 1.07837476e-01 -1.41011715e+00 -7.89942622e-01
1.42692626e-01 -2.94704705e-01 1.59569371e+00 4.64598715e-01
9.56170738e-01 6.89314842e-01 2.93150693e-01 -8.68705451e-01
-5.14801025e-01 -5.44819593e-01 -5.47478139e-01 -1.64304003e-01
-3.34092051e-01 -2.94240594e-01 4.17251550e-02 -3.49578857e-01] | [8.257905006408691, 2.483374834060669] |
dc45eb96-c3eb-4aa5-bbb3-ed7d35971f94 | re-move-an-adaptive-policy-design-approach | 2303.07622 | null | https://arxiv.org/abs/2303.07622v1 | https://arxiv.org/pdf/2303.07622v1.pdf | RE-MOVE: An Adaptive Policy Design Approach for Dynamic Environments via Language-Based Feedback | Reinforcement learning-based policies for continuous control robotic navigation tasks often fail to adapt to changes in the environment during real-time deployment, which may result in catastrophic failures. To address this limitation, we propose a novel approach called RE-MOVE (\textbf{RE}quest help and \textbf{MOVE} on), which uses language-based feedback to adjust trained policies to real-time changes in the environment. In this work, we enable the trained policy to decide \emph{when to ask for feedback} and \emph{how to incorporate feedback into trained policies}. RE-MOVE incorporates epistemic uncertainty to determine the optimal time to request feedback from humans and uses language-based feedback for real-time adaptation. We perform extensive synthetic and real-world evaluations to demonstrate the benefits of our proposed approach in several test-time dynamic navigation scenarios. Our approach enable robots to learn from human feedback and adapt to previously unseen adversarial situations. | ['Dinesh Manocha', 'Amrit Singh Bedi', 'Pratap Tokekar', 'Prithvi Poddar', 'Kasun Weerakoon', 'Souradip Chakraborty'] | 2023-03-14 | null | null | null | null | ['continuous-control'] | ['playing-games'] | [ 1.06037192e-01 5.03112614e-01 2.72991568e-01 -2.40549609e-01
-5.99778950e-01 -8.32549632e-01 5.68558633e-01 -5.82102537e-02
-9.88243043e-01 1.30121386e+00 -1.05826959e-01 -4.07024443e-01
-1.51674166e-01 -7.64138341e-01 -1.05537200e+00 -5.66942930e-01
-3.46174657e-01 5.26697755e-01 5.01731575e-01 -5.72657287e-01
2.57128358e-01 3.82557362e-01 -1.33387005e+00 -3.17228317e-01
8.35255504e-01 5.73696852e-01 4.40089554e-01 9.49690759e-01
3.46267939e-01 8.04327667e-01 -6.83302164e-01 2.65931010e-01
5.57002306e-01 -1.93294585e-01 -6.80749774e-01 -3.04240733e-01
-4.55166519e-01 -7.42984116e-01 -2.80006498e-01 1.01451027e+00
6.34507477e-01 6.71184242e-01 4.97371525e-01 -1.37311482e+00
-3.06612074e-01 5.97149014e-01 -1.85317487e-01 1.94737270e-01
7.32565701e-01 8.31632257e-01 1.29385799e-01 -1.70576423e-01
5.32950401e-01 1.47094834e+00 3.44191283e-01 9.18172896e-01
-9.18466508e-01 -6.33565426e-01 5.66178620e-01 1.57970771e-01
-7.09860444e-01 -3.42323422e-01 4.49723959e-01 -3.97284895e-01
9.77173507e-01 -1.13811694e-01 4.05885935e-01 1.40061963e+00
5.01039743e-01 5.26368678e-01 1.20208538e+00 -3.01593810e-01
7.45323718e-01 -1.33434281e-01 -6.43203080e-01 6.46193922e-01
7.53476098e-03 7.90420294e-01 -3.20067018e-01 -2.93974638e-01
6.06786132e-01 -3.57862651e-01 -2.38381699e-01 -1.84841946e-01
-1.30874968e+00 5.07989824e-01 6.36651695e-01 -3.08239073e-01
-7.18193889e-01 5.62559605e-01 4.42506850e-01 6.58839703e-01
-2.90871203e-01 7.56860435e-01 -4.78117704e-01 -4.87903565e-01
3.18772160e-02 4.96106088e-01 7.77872801e-01 9.44505155e-01
5.70929170e-01 4.47745055e-01 1.92438588e-02 5.45842886e-01
2.79337585e-01 8.91118646e-01 6.76166534e-01 -1.42612326e+00
5.98359287e-01 -1.43755833e-02 8.38691533e-01 -7.44054317e-01
-6.18302524e-01 -1.43227950e-01 -2.14223281e-01 7.97089756e-01
1.39309064e-01 -7.26989210e-01 -1.08960021e+00 1.92383540e+00
5.11056781e-01 5.07697649e-02 3.99492800e-01 1.06653452e+00
-6.36506965e-03 5.09841263e-01 2.14882523e-01 3.95497233e-02
5.89075923e-01 -6.42541409e-01 -5.06466389e-01 -6.85072660e-01
4.33951318e-01 -2.98811913e-01 1.16981733e+00 5.47511995e-01
-6.67348504e-01 -3.00060064e-01 -1.27633286e+00 9.09121990e-01
-1.85059443e-01 -1.70077503e-01 1.97846651e-01 1.75040945e-01
-1.01231408e+00 4.85887349e-01 -1.11297178e+00 -5.88117540e-01
-1.11118853e-01 5.39240718e-01 -3.61657053e-01 4.29013222e-02
-1.40990639e+00 1.07010901e+00 7.45920599e-01 1.67184070e-01
-1.60030544e+00 1.82244733e-01 -8.69773567e-01 -5.09896398e-01
8.77539098e-01 -5.70087254e-01 1.77008116e+00 -6.56367302e-01
-2.21897268e+00 1.36464059e-01 3.28024149e-01 -7.88452804e-01
6.87390745e-01 -5.15033364e-01 -1.30869865e-01 -1.73490383e-02
1.44581318e-01 8.17656755e-01 9.10005391e-01 -1.53680742e+00
-7.33635962e-01 -1.76156670e-01 4.53883499e-01 4.59175617e-01
8.66599567e-03 -5.14084935e-01 -4.77970168e-02 -5.27106106e-01
-1.50828706e-02 -1.39926326e+00 -5.24538636e-01 -2.42582276e-01
-2.86904961e-01 1.24851502e-01 7.14041412e-01 -1.47471815e-01
6.54855788e-01 -1.85261786e+00 1.66796476e-01 6.90434128e-02
-4.71294194e-01 3.08760704e-04 -2.52854317e-01 6.36689067e-01
5.10995507e-01 -1.44783199e-01 -2.47214988e-01 7.40444437e-02
9.16345790e-02 7.87575543e-01 -4.96415854e-01 2.38849819e-01
5.87828010e-02 5.08350968e-01 -1.22866464e+00 -4.22004089e-02
2.61664540e-01 -2.63758097e-02 -6.51773691e-01 4.76246208e-01
-5.39956152e-01 1.05152404e+00 -9.93981421e-01 5.32161474e-01
1.45273611e-01 4.43526030e-01 -6.84081577e-04 5.68940699e-01
2.88719740e-02 -7.74725154e-02 -9.67270195e-01 1.42030752e+00
-6.99421704e-01 2.21511096e-01 3.57757568e-01 -6.78239822e-01
8.37206364e-01 1.80060431e-01 2.49924496e-01 -9.09141839e-01
3.66321146e-01 2.21704617e-01 1.71699421e-03 -6.38732731e-01
4.53851402e-01 1.36063695e-02 -5.23313999e-01 3.45023394e-01
-2.61280239e-01 -5.03417909e-01 -2.99363256e-01 -4.37069535e-02
1.53407705e+00 6.19940579e-01 2.26139367e-01 1.09260648e-01
6.35597110e-01 1.17806941e-01 7.17675567e-01 1.15040040e+00
-7.44163930e-01 6.04285710e-02 2.07065567e-01 -3.06095362e-01
-8.42566431e-01 -9.56718028e-01 5.91865003e-01 1.25126922e+00
5.04819155e-01 2.34822229e-01 -6.42730176e-01 -8.25362921e-01
1.49574935e-01 1.21477044e+00 -8.09937298e-01 -4.91341740e-01
-7.63579369e-01 -2.70062417e-01 4.84715164e-01 3.94137204e-01
4.18744564e-01 -1.62113392e+00 -1.39196920e+00 4.26413536e-01
1.54113006e-02 -8.43711019e-01 -2.86697239e-01 5.71610630e-01
-5.12705028e-01 -9.50890124e-01 -3.69122297e-01 -3.34365845e-01
7.11475432e-01 -1.90340966e-01 4.77615744e-01 -1.56682879e-01
1.19851068e-01 6.77265882e-01 -5.27964652e-01 -4.23035651e-01
-7.88580298e-01 -9.89907831e-02 5.89037776e-01 -5.88068485e-01
-3.61307323e-01 -6.00865662e-01 -5.34770310e-01 6.73699677e-01
-8.40769470e-01 -3.85281563e-01 5.91392934e-01 1.02479959e+00
5.31764209e-01 -2.19792016e-02 8.75913143e-01 -6.71915472e-01
1.24794137e+00 -6.68646455e-01 -7.47678399e-01 1.35551989e-01
-5.70205688e-01 5.33709168e-01 1.02640212e+00 -7.22065091e-01
-1.01029396e+00 5.52924275e-02 -1.86111573e-02 -2.53118336e-01
-1.65210828e-01 4.24985737e-01 1.99921161e-01 1.03926407e-02
9.39122438e-01 2.01736018e-01 -1.83886975e-01 2.83653252e-02
4.47367758e-01 5.96279979e-01 8.91711771e-01 -1.07542300e+00
8.10847938e-01 2.28406519e-01 -2.18535393e-01 -1.09377936e-01
-2.41946220e-01 1.91772759e-01 -2.35145554e-01 -4.92525160e-01
5.98118901e-01 -7.01257169e-01 -8.63288701e-01 4.11634177e-01
-8.17091525e-01 -9.89299238e-01 -2.04928085e-01 4.24651295e-01
-1.09025693e+00 1.18325546e-01 -1.98610842e-01 -1.22131121e+00
-1.72762379e-01 -1.36604369e+00 7.16304243e-01 5.48647761e-01
-3.12463194e-02 -5.52834272e-01 9.24584866e-02 -3.84602658e-02
5.71145952e-01 4.57813799e-01 5.32600939e-01 -5.46416581e-01
-3.91592115e-01 -2.00147733e-01 4.30484861e-01 -1.11599045e-03
4.84589115e-02 -4.74757373e-01 -5.73088944e-01 -7.49926388e-01
-1.65512085e-01 -8.73821199e-01 3.51437122e-01 -2.80218795e-02
5.79448938e-01 -5.93847334e-01 -3.56002331e-01 2.33428001e-01
1.10089052e+00 7.29270339e-01 2.96113968e-01 9.82917011e-01
4.28319052e-02 3.58853877e-01 1.32957888e+00 8.21111977e-01
3.71789485e-01 4.74092603e-01 1.09918356e+00 7.95305550e-01
4.13588613e-01 -5.83333492e-01 8.32306206e-01 2.11926401e-01
2.95031875e-01 -4.69473660e-01 -8.53295028e-01 4.14889097e-01
-2.07575035e+00 -5.86795688e-01 7.17177391e-01 2.33888626e+00
6.96236253e-01 7.16136456e-01 -5.71998097e-02 -3.73758435e-01
5.40426850e-01 -2.13172913e-01 -1.28048813e+00 -3.82883012e-01
3.22682679e-01 -3.66732389e-01 8.96220803e-01 7.40588427e-01
-8.72420371e-01 1.19855583e+00 5.94151306e+00 3.52941334e-01
-1.37068498e+00 9.22190212e-03 9.02732983e-02 1.43381103e-03
-1.75916150e-01 7.34796673e-02 -3.26262623e-01 4.77099746e-01
7.52909601e-01 -1.23334341e-01 7.60442913e-01 1.04951549e+00
3.79241794e-01 -6.40070856e-01 -9.04131413e-01 3.75002772e-01
-3.44062775e-01 -5.04535675e-01 -4.30155545e-01 -2.04029009e-01
4.56286579e-01 2.35054031e-01 1.28379717e-01 8.54183614e-01
1.26602614e+00 -6.90760314e-01 1.00589538e+00 6.46862388e-01
6.52174890e-01 -7.67754614e-01 7.41629183e-01 9.71352756e-01
-7.48567402e-01 -6.10717356e-01 -2.16023147e-01 -1.49515137e-01
3.91206741e-01 -3.03925186e-01 -1.47534311e+00 4.50032532e-01
6.80240452e-01 -3.75039596e-03 -2.33408719e-01 8.41163456e-01
-7.27334559e-01 4.50429678e-01 -5.13005853e-01 -4.18725193e-01
4.41326201e-01 2.20517233e-01 9.48660076e-01 5.73868275e-01
4.00575459e-01 1.73998699e-01 6.79805636e-01 5.26369631e-01
4.84316111e-01 -4.88110006e-01 -7.10658431e-01 1.18861951e-01
8.39835465e-01 6.27530098e-01 -5.45433581e-01 5.22495881e-02
4.32275325e-01 1.05704939e+00 4.46881384e-01 6.09771371e-01
-7.98928797e-01 -4.22611058e-01 4.63552535e-01 -2.14311630e-01
1.40312478e-01 -5.37703931e-01 2.24391252e-01 -5.11183739e-01
-8.95304680e-02 -7.55156815e-01 2.53126830e-01 -9.35765147e-01
-1.04071593e+00 8.00454080e-01 3.98114249e-02 -1.37028730e+00
-6.71753705e-01 -2.59729028e-01 -4.06292140e-01 5.66407263e-01
-1.26849198e+00 -6.85587823e-01 -1.57637581e-01 4.58796382e-01
6.50172889e-01 -4.01063591e-01 7.33269870e-01 -3.16408068e-01
-3.25590223e-01 4.93136704e-01 -6.96918592e-02 -3.44650596e-01
8.20879519e-01 -1.17125499e+00 3.65218103e-01 6.64973259e-01
-7.03078091e-01 4.46688652e-01 1.29059660e+00 -1.00291955e+00
-1.36352253e+00 -9.51302230e-01 -3.12689066e-01 -1.47892743e-01
5.65238357e-01 -1.78523809e-02 -5.39342701e-01 7.04111218e-01
4.55787890e-02 -9.55041274e-02 3.73268165e-02 -4.94823962e-01
-2.11106632e-02 1.04519881e-01 -1.57543623e+00 8.45113635e-01
8.43831360e-01 -1.21181801e-01 -6.07049048e-01 -4.52330895e-02
1.18907726e+00 -7.24707544e-01 -3.42224926e-01 6.67669177e-01
4.54235405e-01 -6.89081192e-01 5.61136305e-01 -6.07399046e-01
-9.32514891e-02 -6.08812571e-01 -2.57501364e-01 -1.81781101e+00
-9.19035971e-02 -9.27296698e-01 4.64707986e-02 6.77523136e-01
4.10915554e-01 -9.39919591e-01 6.16495967e-01 5.69584727e-01
-1.50194317e-01 -6.14770770e-01 -1.14478135e+00 -7.26495445e-01
-5.66796698e-02 -4.17878777e-01 4.51654792e-01 4.00231749e-01
8.48815143e-02 -2.49634102e-01 -6.23376012e-01 5.57801723e-01
3.14286143e-01 -4.86617565e-01 1.01047099e+00 -4.63263005e-01
-4.72999603e-01 -4.13191356e-02 -2.58801311e-01 -1.05104339e+00
2.49126121e-01 -3.17470640e-01 9.94328737e-01 -1.47600091e+00
-5.29696465e-01 -8.40834200e-01 -4.06171799e-01 6.80239975e-01
-1.17076516e-01 -4.55059379e-01 2.58966565e-01 2.30852604e-01
-8.30597222e-01 1.02150273e+00 1.14273131e+00 -1.84398144e-01
-3.90899390e-01 2.64444351e-01 -3.05595368e-01 5.36997437e-01
9.95472789e-01 -4.66586858e-01 -7.02627838e-01 -6.40096307e-01
3.04206729e-01 5.55322230e-01 1.11397304e-01 -1.46586359e+00
4.29196566e-01 -6.66863978e-01 8.87553468e-02 -1.49735674e-01
2.18697444e-01 -9.37428534e-01 -1.27233490e-01 8.19495082e-01
-5.13690472e-01 4.09462482e-01 5.11278510e-01 1.21963358e+00
1.38264418e-01 -2.92075425e-01 5.16849041e-01 -2.90342867e-01
-7.83443153e-01 -1.02093674e-01 -9.79010999e-01 -1.25224516e-01
1.36729634e+00 -3.75613384e-02 -3.29441160e-01 -8.64856005e-01
-8.52589190e-01 1.04715860e+00 5.32586753e-01 7.24726140e-01
6.51136041e-01 -1.00571978e+00 -2.10189193e-01 -2.52794325e-02
5.48388027e-02 1.36145398e-01 1.76021293e-01 3.47742438e-01
-5.93497455e-01 1.85067326e-01 -4.36585456e-01 -3.27347219e-01
-6.46564066e-01 5.81935346e-01 6.16604924e-01 -1.51016498e-02
-3.13404918e-01 9.84571338e-01 -2.24402666e-01 -1.02854741e+00
4.81601864e-01 -1.76700056e-01 -1.06150754e-01 -7.09718704e-01
1.09250210e-01 6.64858967e-02 -2.80344486e-01 -1.39547646e-01
-3.24055731e-01 8.47233608e-02 3.72376889e-02 -8.48055601e-01
9.61211264e-01 -3.60210240e-01 4.02035683e-01 4.39779162e-01
4.00888383e-01 -1.10122800e-01 -1.96834779e+00 -1.29640236e-01
1.47721708e-01 -3.05092007e-01 -3.16481531e-01 -1.33079064e+00
-5.42019725e-01 3.36038053e-01 7.77986526e-01 -1.24102540e-01
8.78929317e-01 -4.29752767e-01 6.79784298e-01 1.16643751e+00
1.20010042e+00 -1.45584702e+00 4.53063726e-01 8.11087251e-01
1.04767525e+00 -1.22517872e+00 -2.51501322e-01 2.18354017e-01
-8.73059392e-01 9.49207664e-01 1.24537206e+00 -1.97496027e-01
3.78986895e-01 2.12112486e-01 4.21409041e-01 1.53136864e-01
-9.13531840e-01 2.97264978e-02 -5.38633108e-01 8.83870900e-01
-5.18535614e-01 9.48465392e-02 -1.57194640e-02 2.69655168e-01
-4.50877488e-01 -7.19096214e-02 7.10661113e-01 1.68564069e+00
-9.11805570e-01 -1.02999973e+00 -6.03635013e-01 1.45530879e-01
9.13727134e-02 6.93068743e-01 -4.14597124e-01 5.56853592e-01
1.16252214e-01 1.06406665e+00 -4.14159656e-01 -6.88361704e-01
3.93591255e-01 -4.66801487e-02 2.97656953e-01 -5.10092795e-01
-3.84753644e-01 -2.40466297e-01 5.80127388e-02 -7.51685560e-01
7.73948133e-02 -5.57682514e-01 -1.91053700e+00 2.81764299e-01
-6.81416094e-02 1.23784378e-01 5.96648395e-01 9.26478088e-01
3.68324786e-01 7.78838694e-01 6.58056319e-01 -9.38957572e-01
-1.25182104e+00 -9.83392715e-01 9.81767029e-02 6.86146468e-02
5.94160616e-01 -1.08852458e+00 -3.45235556e-01 -3.46011788e-01] | [4.398383140563965, 1.6482456922531128] |
4f70ddc5-f801-4a4c-a901-6bcc5e1710e8 | prompt-tuning-based-adapter-for-vision | 2303.15234 | null | https://arxiv.org/abs/2303.15234v1 | https://arxiv.org/pdf/2303.15234v1.pdf | Prompt Tuning based Adapter for Vision-Language Model Adaption | Large pre-trained vision-language (VL) models have shown significant promise in adapting to various downstream tasks. However, fine-tuning the entire network is challenging due to the massive number of model parameters. To address this issue, efficient adaptation methods such as prompt tuning have been proposed. We explore the idea of prompt tuning with multi-task pre-trained initialization and find it can significantly improve model performance. Based on our findings, we introduce a new model, termed Prompt-Adapter, that combines pre-trained prompt tunning with an efficient adaptation network. Our approach beat the state-of-the-art methods in few-shot image classification on the public 11 datasets, especially in settings with limited data instances such as 1 shot, 2 shots, 4 shots, and 8 shots images. Our proposed method demonstrates the promise of combining prompt tuning and parameter-efficient networks for efficient vision-language model adaptation. The code is publicly available at: https://github.com/Jingchensun/prompt_adapter. | ['Changyou Chen', 'Zihao Lin', 'Jiayu Qin', 'Jingchen Sun'] | 2023-03-24 | null | null | null | null | ['few-shot-image-classification'] | ['computer-vision'] | [-2.53490475e-03 -2.87420988e-01 -4.47357893e-01 -3.76660109e-01
-7.71377087e-01 -3.72712553e-01 7.84773171e-01 -2.21520349e-01
-7.12526441e-01 5.24017930e-01 3.86005431e-01 -1.27504066e-01
2.23809153e-01 -3.06714952e-01 -7.17771828e-01 -4.82654691e-01
5.62265098e-01 4.10966486e-01 3.88771772e-01 -1.79635212e-01
1.23408981e-01 8.11654143e-03 -1.36854875e+00 1.73197955e-01
6.84996247e-01 8.32610667e-01 4.90723461e-01 7.60309041e-01
-2.98990667e-01 6.89590216e-01 -2.29323506e-01 -5.20489216e-01
2.98317254e-01 -5.23205996e-01 -6.75796866e-01 -1.12420104e-01
4.75866467e-01 -6.33588195e-01 -2.84130931e-01 7.27332771e-01
9.11262274e-01 4.08881456e-01 6.83648646e-01 -1.31307364e+00
-9.94616687e-01 6.14383578e-01 -6.05862379e-01 4.53758538e-01
-3.55574042e-01 7.09068000e-01 9.32889104e-01 -1.08430171e+00
5.54930508e-01 1.12321007e+00 6.91487074e-01 1.08181381e+00
-1.22120476e+00 -8.48991930e-01 1.71633735e-01 5.72085619e-01
-1.25758600e+00 -9.77466166e-01 4.09236759e-01 -6.84471965e-01
1.37867832e+00 -3.27721328e-01 4.72427666e-01 1.65707815e+00
-1.80319220e-01 8.56365323e-01 5.88052094e-01 -5.29548228e-01
2.30614573e-01 2.43137211e-01 1.06821001e-01 6.37319267e-01
5.50134107e-02 3.69727798e-02 -5.93410313e-01 3.02020516e-02
5.40453017e-01 1.20028883e-01 -2.26346161e-02 -2.57075906e-01
-1.06209421e+00 9.85302448e-01 1.92329943e-01 6.21861294e-02
-3.32057714e-01 5.08912861e-01 7.32301593e-01 8.00467432e-02
6.46427691e-01 4.90021706e-01 -6.14754975e-01 -2.99318910e-01
-8.69708717e-01 -1.06032409e-01 5.45028746e-01 9.64642882e-01
6.30344510e-01 2.16793075e-01 -7.76994765e-01 1.27217162e+00
1.02414116e-01 1.85141936e-01 6.04627252e-01 -1.05412138e+00
3.04410487e-01 2.08242953e-01 -3.48011125e-03 -1.82618573e-01
-2.65064985e-01 -3.52844805e-01 -7.52089083e-01 -3.28463688e-02
2.43502587e-01 -1.46191552e-01 -1.16667235e+00 2.01946807e+00
4.41599041e-01 5.44700205e-01 -9.15912464e-02 6.62420750e-01
8.73131394e-01 5.72870731e-01 5.49166799e-01 3.23883668e-02
1.36944187e+00 -1.61244464e+00 -5.80271840e-01 -3.47627759e-01
5.12316823e-01 -7.55308151e-01 1.70076978e+00 -1.69924766e-01
-8.77478600e-01 -6.69622898e-01 -8.39661896e-01 -2.02953964e-01
-3.00625205e-01 1.87753975e-01 6.47167623e-01 3.14461231e-01
-1.00789142e+00 3.25182080e-01 -8.98198843e-01 -8.68011355e-01
7.43253589e-01 9.17071328e-02 -7.74188042e-02 -1.91319525e-01
-9.91746962e-01 8.56837094e-01 3.16310734e-01 -2.76763678e-01
-1.23909056e+00 -1.05510640e+00 -5.91023862e-01 1.17254622e-01
5.94325244e-01 -1.30191314e+00 1.69577157e+00 -7.84557760e-01
-1.77976418e+00 7.70747602e-01 -1.05504312e-01 -3.93409848e-01
5.62453210e-01 -3.91253352e-01 -7.59567320e-02 -4.49589603e-02
7.02567250e-02 9.44699824e-01 1.08306360e+00 -9.25562263e-01
-4.67695415e-01 -6.04072958e-02 2.51576781e-01 6.84957281e-02
-7.53589809e-01 2.39042237e-01 -8.34820926e-01 -4.63611811e-01
-7.32187212e-01 -8.19540977e-01 -3.76771063e-01 9.14825276e-02
-2.08313346e-01 -2.83524811e-01 4.82277215e-01 -3.82925808e-01
1.35280228e+00 -2.05811191e+00 -1.13727026e-01 -5.55178702e-01
4.15189192e-02 7.61138499e-01 -6.23977721e-01 5.04130423e-01
1.85785443e-01 9.70951542e-02 -1.08184472e-01 -7.40175605e-01
3.57467984e-03 1.92100167e-01 -9.72410962e-02 1.06209889e-01
1.53916329e-01 1.14547706e+00 -8.93370867e-01 -4.45258141e-01
2.03955740e-01 5.49670935e-01 -5.89072943e-01 4.27683443e-01
-3.81727099e-01 4.80175138e-01 -3.57222259e-01 5.44911265e-01
2.98532069e-01 -6.25003338e-01 -4.38891381e-01 -1.05920911e-01
2.48563522e-03 -4.97140102e-02 -7.38679290e-01 1.89655900e+00
-5.06443739e-01 5.73997378e-01 -2.08535194e-01 -7.66252398e-01
5.75797021e-01 3.31743479e-01 1.87660217e-01 -6.42095387e-01
1.10285871e-01 -1.29480720e-01 -2.49383315e-01 -8.34756017e-01
3.15824836e-01 4.55880985e-02 2.06767261e-01 4.66830343e-01
4.51384485e-01 3.62489112e-02 2.90523380e-01 3.05822134e-01
1.21497917e+00 2.10553065e-01 4.67442840e-01 1.23521350e-01
2.26703063e-01 -6.73130974e-02 5.33408582e-01 1.07391846e+00
-6.07781351e-01 5.71537256e-01 9.60438251e-02 -3.53809208e-01
-1.25535023e+00 -7.85904348e-01 2.13017955e-01 1.79091728e+00
-2.78146088e-01 -5.11815846e-01 -7.83181429e-01 -5.21302283e-01
-1.28927961e-01 8.46338332e-01 -7.38598168e-01 -4.20207590e-01
-1.14589326e-01 -6.73852801e-01 5.71601272e-01 8.15354466e-01
5.15820622e-01 -1.07134354e+00 -5.48329532e-01 2.20641300e-01
-1.89191177e-01 -1.27746320e+00 -9.03000891e-01 1.29978240e-01
-5.54233074e-01 -7.45508373e-01 -8.68812978e-01 -5.48875511e-01
5.19351900e-01 5.44118106e-01 1.17273676e+00 1.37785189e-02
-3.41049075e-01 6.54855728e-01 -4.94631439e-01 -5.49039960e-01
-3.46199423e-01 4.33070600e-01 1.72182962e-01 -4.47302870e-02
4.91964906e-01 -5.17789483e-01 -7.93178678e-01 2.56907940e-01
-8.94774079e-01 3.81569862e-01 5.81881702e-01 1.01353168e+00
5.33737600e-01 -7.85143256e-01 7.32056081e-01 -9.89416778e-01
6.78628802e-01 -6.09252274e-01 -5.30650616e-01 5.22990346e-01
-8.09770942e-01 1.05024762e-01 5.87214470e-01 -8.17276239e-01
-1.21596432e+00 9.88886431e-02 -1.30579650e-01 -7.47926414e-01
-2.23652735e-01 2.68092990e-01 2.65654504e-01 -9.67235044e-02
9.76751268e-01 -8.70919600e-03 -2.22428471e-01 -5.93576849e-01
7.17232645e-01 7.06745267e-01 5.22534668e-01 -4.61924553e-01
8.11369836e-01 2.94459313e-01 -3.71392757e-01 -6.89438164e-01
-1.12511039e+00 -7.52947927e-01 -6.67659402e-01 -8.91517177e-02
9.39798653e-01 -1.23154211e+00 -1.98717654e-01 5.27968884e-01
-1.05207753e+00 -9.05721009e-01 -3.05676073e-01 4.25604373e-01
-6.18094146e-01 1.76471204e-01 -6.51476681e-01 -6.75194621e-01
-7.61089027e-01 -9.02829409e-01 8.22052896e-01 4.26995218e-01
-2.22211897e-01 -9.22961295e-01 3.85973930e-01 4.50423956e-01
8.50418448e-01 -2.46373802e-01 7.99453497e-01 -6.61101639e-01
-4.12144095e-01 2.95788404e-02 -3.33988935e-01 4.09677744e-01
-1.12851538e-01 5.56295216e-02 -1.13367438e+00 -2.61307150e-01
-3.34414154e-01 -8.18850458e-01 1.22221136e+00 5.94495296e-01
1.16064668e+00 -1.69601321e-01 -2.22565621e-01 1.02614975e+00
1.47783840e+00 -4.82340492e-02 2.87210226e-01 4.56981063e-01
8.65438461e-01 1.47948429e-01 4.84305710e-01 7.33943760e-01
5.64117491e-01 7.38623619e-01 2.35845044e-01 6.64524548e-03
-4.10056710e-01 -2.82925367e-01 3.01684946e-01 6.88426018e-01
-1.49353012e-01 -2.41171315e-01 -9.29520667e-01 6.68515682e-01
-2.03683424e+00 -8.72977972e-01 3.31779718e-01 1.98072183e+00
8.58599603e-01 -1.15002990e-01 2.15719521e-01 -6.89901173e-01
6.61781788e-01 1.64316073e-01 -9.82569277e-01 -3.51825088e-01
1.24497637e-01 -2.64222529e-02 4.37479824e-01 4.02759641e-01
-9.11353946e-01 1.43041706e+00 5.72534132e+00 9.43468690e-01
-9.40821707e-01 7.30105996e-01 5.34255266e-01 -6.16728723e-01
-1.36059254e-01 2.69768648e-02 -1.19938600e+00 4.31353956e-01
1.09208977e+00 -2.34420076e-01 6.63384378e-01 9.91797507e-01
3.00628453e-01 2.54302789e-02 -1.05766284e+00 1.00870204e+00
2.42896438e-01 -1.41182792e+00 1.69173181e-01 -2.04179868e-01
9.27227199e-01 6.47221565e-01 3.69678959e-02 7.50271022e-01
3.91675621e-01 -6.83949232e-01 5.28014541e-01 4.87102062e-01
1.03649545e+00 -3.27048510e-01 4.38128114e-01 3.33823472e-01
-1.09349716e+00 -2.57036835e-01 -6.41356826e-01 1.70899574e-02
2.23777160e-01 2.18381256e-01 -9.49259698e-01 1.25995316e-02
7.65994489e-01 6.25422776e-01 -7.60457754e-01 1.29831183e+00
-2.30744317e-01 8.55861545e-01 -1.16935052e-01 1.23913430e-01
3.34311038e-01 6.88659623e-02 3.31831634e-01 1.41682255e+00
3.34674567e-01 1.83001339e-01 3.67686362e-03 7.19872415e-01
-1.74057037e-01 1.24695189e-01 -5.46013951e-01 -2.14402437e-01
5.98876238e-01 1.28916502e+00 -5.12947857e-01 -5.56609869e-01
-7.33952105e-01 9.78081226e-01 7.56908774e-01 5.87780714e-01
-9.14659560e-01 -1.37253940e-01 7.41134703e-01 -1.56691089e-01
3.93102139e-01 -1.40210018e-01 -2.34570578e-02 -1.23316014e+00
-1.93216562e-01 -5.98727345e-01 5.24972200e-01 -8.06681752e-01
-1.45339108e+00 4.66538846e-01 9.22159255e-02 -9.78556454e-01
-3.25479239e-01 -3.77840519e-01 -7.99062312e-01 5.30148268e-01
-1.62129104e+00 -1.37993503e+00 -5.90586126e-01 7.32677996e-01
1.05781841e+00 -4.31897700e-01 8.33322406e-01 2.71706790e-01
-9.81249750e-01 9.78266418e-01 2.45906249e-01 -1.24226347e-01
1.12101364e+00 -9.29619789e-01 6.45237207e-01 8.84260416e-01
4.18581776e-02 3.75434637e-01 6.40611410e-01 -3.23427320e-01
-1.09002101e+00 -1.20268023e+00 7.14515150e-01 -6.03235781e-01
6.96223557e-01 -3.22051167e-01 -9.73257482e-01 8.03035021e-01
4.37678188e-01 2.03254223e-01 8.26935709e-01 1.22501865e-01
-5.57393968e-01 -1.71918064e-01 -7.55754650e-01 7.20461428e-01
1.39997411e+00 -5.08451045e-01 -2.97528267e-01 4.86524165e-01
9.63180542e-01 -2.02442691e-01 -5.48812211e-01 2.88551182e-01
5.33429325e-01 -7.19109118e-01 1.08819818e+00 -7.77852356e-01
4.84720469e-01 5.11493459e-02 -1.15979359e-01 -1.43526947e+00
-7.50664473e-01 -6.69070184e-01 -4.19860333e-01 1.39042020e+00
4.73744839e-01 -5.35939813e-01 5.04341006e-01 7.65085220e-01
-2.75063455e-01 -7.86080301e-01 -6.09924734e-01 -9.27996755e-01
-1.91330239e-01 -3.63011062e-01 4.51378256e-01 8.31704021e-01
-3.76052201e-01 8.59856188e-01 -7.08144546e-01 -1.75228462e-01
4.79240119e-01 -2.69980967e-01 1.01640093e+00 -9.91847515e-01
-6.41671717e-01 -5.37569821e-01 5.40680364e-02 -8.18627238e-01
2.33611077e-01 -8.00876975e-01 5.33095933e-02 -1.71763861e+00
8.02207351e-01 -2.10929438e-01 -6.21999681e-01 9.55386639e-01
-5.29764831e-01 2.02647641e-01 3.38511020e-01 3.64235222e-01
-9.80336845e-01 7.77054071e-01 1.03039908e+00 -1.52536109e-01
-3.15241843e-01 3.57720219e-02 -8.10314715e-01 5.72446585e-01
8.79586399e-01 -5.30070543e-01 -6.06100738e-01 -8.01796436e-01
-9.14021209e-03 -4.33652103e-01 2.03200027e-01 -9.31258857e-01
4.09492731e-01 -3.12536180e-01 3.37495714e-01 -1.62071109e-01
4.68806565e-01 -4.76039350e-01 9.54979062e-02 2.98234284e-01
-5.20523965e-01 -4.24191616e-02 3.01743567e-01 6.26194239e-01
1.93321586e-01 -3.65158081e-01 1.00450754e+00 -3.03524435e-01
-1.15680277e+00 6.78241611e-01 -4.67115119e-02 3.02604616e-01
9.97515321e-01 -4.27517854e-02 -6.01458192e-01 -3.47308755e-01
-5.01122117e-01 2.95646101e-01 3.94063145e-01 6.42288089e-01
4.55550492e-01 -1.26941037e+00 -7.87145615e-01 -1.29249468e-02
6.08837426e-01 -3.26357871e-01 4.41387981e-01 8.96518171e-01
-4.73282300e-02 2.88124532e-01 -2.91580141e-01 -5.37450552e-01
-1.19128811e+00 5.72717011e-01 1.73616081e-01 -1.77865654e-01
-5.85093856e-01 1.28252709e+00 4.30474073e-01 -2.09886536e-01
3.84546846e-01 1.41013972e-02 -1.33078679e-01 1.47675067e-01
6.69947088e-01 2.92172790e-01 -2.88826346e-01 -3.56995851e-01
-1.23944022e-01 5.69452107e-01 -3.02244574e-01 -7.26724789e-02
1.37120759e+00 -3.74683589e-01 2.94616133e-01 6.15274131e-01
1.06565571e+00 -5.80012798e-01 -1.83503902e+00 -6.76058590e-01
-2.74610579e-01 -4.75916773e-01 3.01398188e-02 -8.39341223e-01
-8.95147264e-01 9.35247600e-01 6.21318817e-01 -3.77728969e-01
1.07939160e+00 6.83764815e-02 8.90525043e-01 5.62441230e-01
2.31207609e-01 -1.34297144e+00 4.26677555e-01 7.72437930e-01
6.46517515e-01 -1.77457047e+00 -2.81615674e-01 2.52871066e-01
-9.51269329e-01 6.72952712e-01 1.04713750e+00 2.77872741e-01
4.50548470e-01 -3.09876576e-02 1.98438153e-01 4.87896837e-02
-1.23719704e+00 -4.40245509e-01 1.93330556e-01 5.99207103e-01
2.12609574e-01 -6.91469908e-02 5.48065864e-02 5.85413337e-01
1.95125774e-01 2.92416334e-01 2.29891554e-01 5.88223517e-01
-5.00587225e-01 -9.36291575e-01 4.37853448e-02 5.37210524e-01
-3.67635816e-01 -3.86217088e-01 -1.48021817e-01 5.45846403e-01
-5.19411229e-02 8.32075655e-01 -1.73968539e-01 -3.16703677e-01
3.23504686e-01 3.60258698e-01 4.33738679e-01 -8.33953440e-01
-4.84185785e-01 -1.36931896e-01 1.09644458e-02 -6.70452237e-01
-2.51514405e-01 -3.62355620e-01 -7.08758175e-01 -2.44803727e-01
-2.04364672e-01 -2.99956322e-01 6.85169518e-01 9.23550129e-01
6.86616123e-01 5.05544186e-01 3.34829897e-01 -1.01843679e+00
-8.02111089e-01 -1.23128736e+00 -1.78265378e-01 3.76745075e-01
3.32019538e-01 -6.62084222e-01 -3.22349876e-01 3.00914913e-01] | [10.088028907775879, 2.2586004734039307] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.